Command and Control

Balanced and Unbalanced

During the summer of 1979, James L. “Skip” Rutherford III was working in the Little Rock office of Senator David H. Pryor. Rutherford was twenty-nine years old. He’d grown up in Batesville, Arkansas, a small town in the northern part of the state, attended the University of Arkansas, and edited the student newspaper there. After graduation he did public relations work for a bank in Fayetteville. The job introduced him to Pryor, who was running for a seat in the United States Senate, after two terms as the governor of Arkansas. Pryor was a new breed of southern Democrat, an opponent of racism and segregation, a supporter of women’s rights, a progressive who greatly enjoyed meeting with voters, rich or poor, in every corner of the state. Rutherford worked as a volunteer for the campaign and joined Pryor’s staff after the election, representing the senator at events throughout Arkansas. And then one day Rutherford took a call from someone at Little Rock Air Force Base, a young airman who wanted to meet with Pryor confidentially. The airman sounded nervous. When Rutherford asked what this was about, the airman said: “It’s about the Titan missiles.”

Skip Rutherford didn’t consider himself an expert on intercontinental ballistic missiles. But he’d served in the Arkansas Air National Guard for six years, spending one weekend a month at Little Rock Air Force Base. He knew a lot of people at the base and felt comfortable there. The airman agreed to meet with Rutherford at the federal building in Little Rock, after hours, to avoid being seen — and brought a couple of other guys who worked with the Titan II. They were about Rutherford’s age. They didn’t want their names used as the source of any information. They were scared about getting into trouble. And most of all they were scared about what was happening at the Titan II silos in Arkansas.

The missiles were old, the airmen said, and most of them leaked. The portable vapor detectors and the vapor detectors in the silos often didn’t work. Spare parts were hard to find. The Propellant Transfer System crews were overworked, sometimes spending fifteen or sixteen hours on the job. And many of the young PTS technicians weren’t adequately trained for the tasks they were being ordered to perform. After that first meeting, Rutherford secretly met with other airmen from the base and took their calls from pay phones late at night. He spoke to roughly a dozen members of the 308th Strategic Missile Wing, promising not to reveal their identities to the Air Force. And they all said basically the same thing: the Titan II was a disaster waiting to happen.

Rutherford told Senator Pryor about the meetings. Pryor was disturbed by the information and decided that something had to be done. He wrote to Dr. Hans S. Mark, the secretary of the Air Force, asking for details about the staff shortages and training deficiencies at Little Rock Air Force Base. And Pryor learned that other members of Congress were concerned about the Titan II. Representative Dan Glickman, a Democrat, and Senator Bob Dole, a Republican, had already asked the Air Force to launch a formal investigation of safety problems with the Titan II. Glickman and Dole were both from Kansas, where some of the missile’s flaws had been revealed during an accident the previous summer.

AT LAUNCH COMPLEX 533-7, about an hour southeast of Wichita, Kansas, the final stages of a missile recycle were being completed. A Titan II had been removed from 3–7 and returned to McConnell Air Force Base, where it would undergo routine maintenance checks. A replacement missile had been lowered into the silo. On the morning of August 24, 1978, a PTS crew arrived at the complex to pump oxidizer into the tanks. The fuel would be added the following day, and then the warhead would be placed atop the Titan II, finishing the recycle. On the main floor of the control center, the head of the PTS crew, Staff Sergeant Robert J. Thomas, briefed the missile combat crew commander, First Lieutenant Keith E. Matthews, about the work that would be done that day. A trainee, Airman Mirl Linthicum, would be acting as PTS team chief, supervising the procedure from the control trailer topside.

Oxidizer lines were attached to the stage 1 and stage 2 tanks, and both were full in about an hour. The lines were thick, heavy hoses through which the propellant flowed. Airman Erby Hepstall and Airman Carl Malinger put on RFHCO suits and entered the silo to disconnect the lines. Malinger had never been inside a Titan II silo before. He was nineteen years old and new to the Air Force, accompanying Hepstall that day for on-the-job training. The removal of the stage 2 lines, near the top of the missile, went smoothly. Hepstall and Malinger rode the elevator down to disconnect the lines from stage 1. Standing on a platform near the bottom of the missile, they unscrewed one of them. A powerful stream of oxidizer, like water suddenly released from a fire hydrant, hit Malinger’s chest and the faceplate of his helmet and knocked him down. Hepstall tried to reconnect the line, but it wouldn’t screw back on. Oxidizer poured from the missile, fell into the W below it, and then rose as a thick, reddish brown cloud of vapor.

Inside the top level of the control center, Lieutenant Matthews was preparing his lunch when a Klaxon sounded. Down below, the deputy commander, Second Lieutenant Charles B. Frost, sat at the launch control console. Frost wore a headset and monitored the PTS team on the radio. He pushed a button on the console and turned off the Klaxon, assuming that a puff of oxidizer had set it off when the lines were disconnected. That happened all the time. The Klaxon sounded again, and Frost heard screams over the radio.

“Oh my God, the poppet.”

“What was the poppet?” Frost said into his headset. “What’s wrong?”

Matthews came down the stairs as warning lights flashed on the console: OXI VAPOR LAUNCH DUCT, VAPOR SILO EQUIP. AREA, VAPOR OXI PUMP ROOM.

“Get out of here, let’s get out,” a voice yelled over the radio.

“Where are you?” Frost asked. The sounds on the radio were chaotic. People were talking at the same time, they were shouting and screaming and drowning one another out. Frost pushed the override button, blocking everybody else’s radio transmission, and ordered: “Come back to the control center.”

“I can’t see,” somebody said.

Lieutenant Matthews walked over to the blast door protecting the control center. He tried to open the door and see what was going on. The blast door wouldn’t open. And Matthews got a whiff of something that smelled a lot like Clorox bleach. It smelled like oxidizer.

In the control trailer topside, Airman Linthicum, the trainee running his first recycle, heard the shouting on the radio but couldn’t understand what was being said. Linthicum ran out of the trailer, trying to get better reception on a portable headset, and saw a reddish cloud rising from the exhaust vents. Another member of the PTS crew left the trailer, found Sergeant Thomas — the most experienced technician at the site — and told him something had gone wrong. Thomas was twenty-nine years old. He saw the oxidizer, ran to the access portal, and asked the control center for permission to enter the launch complex.

Lieutenant Frost granted the permission, unlocking the outer steel door for Thomas and then the door at the bottom of the entrapment area. All the hazard lights on Frost’s console seemed to be flashing at once, including FUEL VAPOR LAUNCH DUCT, which made no sense. Frost kept asking the PTS team chief where they were in the checklist when the accident happened, hoping to find the right emergency checklist for dealing with it. But the radio still didn’t work properly. Frost pulled out different tech manuals and flipped through their pages. He wasn’t sure what they were supposed to do.

“Hey, I smell Clorox,” Matthews said. He told the missile crew to set up a portable vapor detector in front of the door, to close the blast valve and the blast damper, protecting the air supply of the control center.

The missile facilities technician, Senior Airman Glen H. Wessel, placed a vapor detector near the blast door. He could smell oxidizer. The detector quickly registered one to three parts per million; somehow the stuff was getting into the control center. Wessel told his commander that the room was being contaminated with oxidizer. They both tried to open the blast door, but it wouldn’t budge. The crew was locked inside the control center.

The two PTS technicians waiting in the blast lock, serving as backup, had no idea what was happening in the silo. They could hear screams on the radio, but nobody would answer them. And then the door from the long cableway suddenly swung open, and Hepstall appeared. Oxidizer had turned the faceplate of his helmet white. It was so opaque you couldn’t see his face.

Hepstall pulled off the helmet. He was sobbing. He said Malinger’s still down there, we have to go and get him out. If anything happens to Malinger, he said, I’ll never forgive myself.

Hepstall had left his trainee in the silo, amid a thick cloud of oxidizer, found his way to the elevator, and ridden it five levels to the long cableway.

The door to the blast lock opened, and Sergeant Thomas walked inside. He saw Hepstall sobbing, heard that Malinger was missing, and put on one of the backup team’s RFHCO suits. Without a moment’s hesitation, Thomas had decided to search for Malinger.

Hepstall offered to go with him and grabbed a fresh helmet. Wearing the RFHCOs, they opened the door and headed down the long cableway toward the silo. The air was becoming thick with oxidizer.

The PTS backup team waited anxiously in the blast lock. Moments later, the door swung open. Hepstall stumbled inside and fell to the ground coughing. He hadn’t made it very far. The new helmet leaked, and oxidizer was getting into his RFHCO. Hepstall took off the suit, got into another one, and left for the silo again.

On the bottom floor of the control center, Wessel was amazed by how hard it was to open the escape hatch. The ratchet that you needed to use felt really heavy. He and the ballistic missile analyst technician, Danford M. Wong, took turns with it, wearing their gas masks. They were highly motivated. The blast door still wouldn’t open, and this looked like their only way out.

Lieutenant Frost was still attempting, without success, to reach the PTS team in the silo, Sergeant Thomas, and the PTS guys in the trailer, using the telephone and the radio. It wasn’t easy with a gas mask on. Frost would pull off the mask momentarily, speak, put the mask back on, and listen for some response. Nobody answered him. And then, clear as a bell, he heard Malinger shouting over the radio.

“My God, help us, help us, we need help.”

“Hey, door eight is locked, we’re locked in, you guys get out,” Frost told him.

Malinger kept repeating that he needed help, and Frost tried to make him understand that the blast door was stuck.

The emergency phone rang, and Frost answered it. Someone was outside blast door 8, asking for help.

“Hey, you guys, get out of here, get out of here now,” Frost said, “just get out, door eight is locked, so you guys get out.”

Wessel and Wong could hear the commotion on the floor above them and cranked the ratchet on the escape hatch as fast they could.

Blast door 8 swung open, and Malinger ran into the control center, carrying his helmet, yelling that Sergeant Thomas was dead. A cloud of oxidizer followed him, and then Hepstall came in, without a helmet, and collapsed onto the floor. He landed near the stairs, as Malinger kept screaming. None of it made sense to the missile crew.

Commander Matthews said, “Come help me,” to Frost, and they entered the blast lock. Sergeant Thomas lay unconscious on the floor. They picked him up, carried him into the control center, and shut the door. Thomas was having convulsions, his head nodding side to side in the RFHCO helmet. Malinger took off the helmet and started to give him mouth-to-mouth.

“This is three-seven,” Frost told the command post at McConnell Air Force Base. “The locks are on the safe and the keys are in it. We got one man possibly down and we’re evacuating now.”

Thomas died on the floor, staring at the ceiling.

“Where’s the dep, where’s the dep?” Wessel shouted, calling for Frost, their deputy commander. They were getting tired, and they needed his help to open the escape hatch. The light grew dimmer as the control center filled with oxidizer.

Malinger didn’t want to leave Thomas behind. It seemed wrong. After getting knocked down by the powerful stream of oxidizer, Malinger had gotten lost in the silo, near the base of the missile, unable to see more than a few feet, unaware that Hepstall had taken the elevator and left him down there. Sergeant Thomas had found him and brought him out, and now Malinger didn’t want to leave Thomas on the floor.

“We’ll get him later,” Frost said, heading downstairs to work on the hatch.

Matthews helped Malinger and Hepstall down the stairs and then helped them take off their RFHCO suits. They said their skin was burning. “My God, please help me,” Hepstall said. “It’s in here with me, it’s with me.”

Matthews went back upstairs and checked the other two levels of the control center, looking for stragglers, just in case. The cloud of oxidizer was now so thick that he couldn’t see more than two or three feet ahead.

The escape hatch was open, finally. Wessel went into it first, crawled through the tunnel, and climbed the ladder as fast as he could. It felt like climbing up a chimney full of smoke, as the oxidizer filled the narrow air shaft. At the top, Wessel pulled the pins and then pushed the metal grating open with his head. Wong was right behind him, and then Frost, who’d paused every few rungs to pull Hepstall up the ladder. Frost wanted to help him — and didn’t want him falling onto Malinger. Lieutenant Matthews went last, closing the hatch behind him to trap the oxidizer in the control center.

The missile crew carried the two injured PTS technicians to the emergency showers on the hardstand to rinse them off. The showers didn’t work.

“Get them under the fire hydrant,” Matthews said.

The crew put Hepstall and Malinger in front of the hydrant and turned it on. Water poured out, and then after a few seconds the hydrant sputtered air and quit. They had to get these men rinsed off, immediately. But the gate to the launch complex was locked. No one had remembered to unlock it before abandoning the control center, and the trucks were parked on the other side of the fence. With help from some of the PTS technicians, the missile crew carried Hepstall and Malinger through the breakaway panel in the fence and placed them in the bed of a pickup.

The crew drove to a nearby farmhouse, and warned the occupants that deadly fumes were rising from the silo. Wong said to leave the area at once — and Frost asked to use their phone. Wessel found a garden hose in the backyard. After spraying the two airmen with water, they drove Hepstall and Malinger to the nearest hospital.

A cloud of oxidizer floated from the launch complex, extending for about a mile and drifting toward the town of Rock, Kansas. The cloud looked like a dark, ominous thunderhead. Local residents didn’t know what it was, and the cars and trucks on Highway 77 drove right through it. Air Force security police soon evacuated the roughly two hundred inhabitants of Rock.

Sergeant Thomas had been left behind, and none of the PTS crew members felt right about that. Even though he was gone, they thought, he shouldn’t be lying down there, alone. Two men volunteered to get him: Mirl Linthicum, the team chief trainee, and Airman John G. Korzenko. They returned to the launch complex and put on RFHCO suits. Linthicum climbed into the escape hatch first, followed by Korzenko. Within seconds, Korzenko had climbed out; oxidizer was leaking into his suit. Linthicum came back moments later; he wasn’t getting enough air in his helmet.

Another PTS team arrived from McConnell, with fresh RFHCO suits and air packs. They wanted to get Thomas out, too. Airman Middland R. Jackson put on a RFHCO and climbed into the escape hatch. He came right back; his helmet was leaking. Jackson grabbed another helmet, tried the escape hatch again, and climbed the ladder all the way to the bottom in his RFHCO. But he’d never been in the escape hatch before, and the oxidizer was so thick down there he couldn’t find the entrance to the control center. He climbed back up the ladder, frustrated and yet determined not to quit.

A few minutes later, Jackson and two other PTS technicians in RFHCOs, Technical Sergeant John C. Mock and Airman Michael L. Greenwell, tried to enter the control center through the access portal. They wandered underground through dense clouds of oxidizer, literally feeling their way down the stairs and through blast doors. They could not see more than a foot or so ahead — and had to stick together because none of their radios worked. They made it to the control center, found Sergeant Thomas on the floor, and carried his body onto the elevator. But no matter how many times they pushed the buttons, the elevator wouldn’t work. They decided to carry Thomas up the stairs. His body was heavy, their suits felt heavy, and it was hot down there. After a few minutes, they couldn’t carry him any farther and had to leave his body on the stairs. Two more PTS technicians in RFHCOs, Sergeant James Romig and Airman Gregory W. Anderson, went down and carried him, then had to quit, because of the heat. The five men took turns going into the complex and carrying Thomas as far as they could. As soon as one group got tired, the other would step in. It took two hours to get Thomas up the stairs and out of the complex.

An investigation of the accident later found the cause of the leak. Someone hadn’t put a filter inside the oxidizer line. But the small rubber O-ring designed to hold the filter had been left inside the line. The O-ring blocked the poppet valve from closing fully, allowing oxidizer to pour out. Nobody accepted responsibility for failing to insert the filter. Oxidizer flowed more quickly without a filter in place — and someone may have deliberately omitted the filter to save time and load the tank quickly.

The blast door leading to the control center wouldn’t open because someone had propped open the blast door across from it with a bungee cord — and both doors couldn’t be open at the same time. Hepstall had used the manual override to unlock blast door 8, and by entering the control center, he’d contaminated it with oxidizer.

Robert J. Thomas was killed by a leak in his RFHCO, most likely at the spot where it intersected with the left glove. Oxidizer may have poured into the suit as he tried to reconnect the line to the missile. The Air Force recommended, in the future, that black vinyl electrical tape be used to seal the interface between the glove and the RFHCO suit more securely. Thomas left a widow and two young sons.

Erby Hepstall died a week and a half later, at the age of twenty-two, his lungs destroyed by oxidizer. His son had just turned two. A small tear in the left leg of Hepstall’s RFHCO suit, about seven eighths of an inch long, had allowed oxidizer to enter it.

Carl Malinger had a stroke, went into a coma, suffered lung and kidney damage, lost the use of his left arm, and spent the next several months in the hospital. He’d enlisted to get training as an automobile mechanic, and his mother later felt enormous anger at the Air Force. Its report on the accident said that Hepstall and Malinger had failed to “comply with [Technical Order] 21M-LGM25C-2-12 which states ‘if disconnect starts to leak… screw disconnect to fully connected position immediately.’” The report suggested that Malinger — never trained for the task and working in a Titan II silo for the first time — was somehow to blame for what happened.

General Curtis LeMay had created an institutional culture at the Strategic Air Command that showed absolutely no tolerance for mistakes. People were held accountable not only for their behavior but for their bad luck. “To err is human,” everyone at the command had been told, “to forgive is not SAC policy.”

BLAMING YOUNG ENLISTED MEN for the accident at Rock, Kansas, didn’t eliminate the problems with the Titan II. The Pentagon had announced in 1967 that the Titan II was no longer needed and would be decommissioned, with the first missiles coming off alert in 1971. But every year the Air Force successfully battled to keep the Titan II. Its warhead was more than seven times more powerful than the warhead carried by the Minuteman II. The United States had about one thousand land-based missiles — and the fifty-four Titan IIs represented roughly one third of their total explosive force. SAC didn’t want to lose all that megatonnage without getting new weapons to replace it. As the Titan II aged, however, its ability to reach the Soviet Union became more uncertain. The last test-launch of a Titan II occurred in 1976, and no more were planned, due to a shortage of missiles and parts.

When Senator Pryor and Skip Rutherford visited a Titan II site in Arkansas, the place looked impressive. But one of Rutherford’s confidential sources later told him that there’d been an oxidizer leak at a nearby launch complex that day — and that the vapor detectors in thirteen of the state’s eighteen silos were broken. Pryor came up with a relatively inexpensive plan for protecting rural communities from fuel and oxidizer leaks at Titan II missile sites: install a siren, at every complex, that would blare whenever the crew turned on the red warning beacon topside. The siren could easily be mounted on the same pole. It would warn neighboring homes and farms of a leak. The Air Force opposed the idea, arguing that a siren “might cause people to leave areas of safety and evacuate into or through areas containing propellant fumes.” Colonel Richard D. Osborn told Pryor that during those rare occasions when civilians needed to be alerted, the combined efforts of Air Force personnel and local law enforcement officers would ensure public safety. Pryor nevertheless decided to seek funding for the sirens through an amendment to a Senate bill.

The Titan II missile wasn’t the only Air Force weapon system having maintenance problems. Amid the defense cutbacks following the Vietnam War, the purchase of new planes and missiles had a much higher priority than buying spare parts for the old ones. During the late 1970s, on a typical day, anywhere from one half to two thirds of the Air Force’s F-15 fighters were grounded for mechanical reasons. The Strategic Air Command had lost more than half of its personnel since 1961. Some of its B-52 bombers were twenty-five years old. And SAC’s aura of invincibility had taken a beating. The highest-ranking officers in the Air Force tended to be “bomber generals” who’d risen through the ranks at SAC — and many of the pilots who flew bombing missions in Vietnam resented their insistence on rigid, centralized control. Tactics designed for executing the SIOP proved ineffective during combat in Vietnam, where the targets were often mobile and flying in a rigid formation could get you shot down. American pilots began to disobey orders, ignore their designated targets, bomb those that seemed more urgent, and lie about it in their reports.

Chuck Horner — who flew more than a hundred missions in Vietnam and later commanded the U.S. and allied air campaign during the first Gulf War — resented the inflexible, “parent-child relationship” that SAC’s bomber generals often demanded. He felt a tremendous anger, shared by many other young officers, about how the Air Force leadership had behaved during the Vietnam War:

Horner vowed that he would “never again be a part of something so insane and foolish.” After the war, thousands of young officers left the Air Force, profoundly disillusioned. Many of those who stayed were determined to change things. And the influence of the Strategic Air Command gradually diminished, as a younger generation of “fighter generals,” who rejected centralization and standardization and rigid planning, who had firsthand experience in real combat and little interest in abstract theories about nuclear war, rose to power.

During the years following the Vietnam War, antimilitary sentiment in the United States became stronger, perhaps, than at any other time in the nation’s history. Vietnam veterans were routinely depicted in books and films as racists, stoners, nutcases, and baby killers. Morale throughout the armed services suffered — and illegal drug use soared. By 1980, according to the Pentagon’s own surveys, about 27 percent of all military personnel were using illegal drugs at least once a month. Marijuana was by far the most popular drug, although heroin, cocaine, and LSD were being used, too. Among the armed services, the Marines had the highest rate of drug use: about 36 percent regularly smoked pot. About 32 percent of Navy personnel used marijuana at least once a month; the proportion of Army personnel was about 28 percent. The Air Force had the lowest rate, about 14 percent. It also had the most powerful warheads and bombs. The surveys by the Department of Defense most likely understated the actual amount of drug use. Random urine tests of more than two thousand sailors at naval bases in Norfolk, Virginia, and San Diego, California, found that almost half had recently smoked pot. Although nuclear weapons and marijuana had recently become controversial subjects in American society, inspiring angry debates between liberals and conservatives, nobody argued that the two were a good combination.

Donald Meyer served as a corporal with the 74th United States Field Artillery Detachment in Germany during the early 1970s. His detachment kept Pershing missiles on alert, ready to fire within fifteen minutes. Each missile carried an atomic warhead ten to twenty times more powerful than the bomb that destroyed Hiroshima. Meyer told the Milwaukee Journal that almost every one of the more than two hundred men in his unit regularly smoked hashish. They were often high while handling secret documents and nuclear warheads. A survey found that one out of every twelve members of the United States Army in Germany was smoking hashish every day. “You get to know what you can handle,” Meyer said. “Too much hash and you would ruin a good thing.”

At Homestead Air Force Base in Florida, thirty-five members of an Army unit were arrested for using and selling marijuana and LSD. The unit controlled the Nike Hercules antiaircraft missiles on the base, along with their nuclear warheads. The drug use at Homestead was suspected after a fully armed Russian MiG-17 fighter plane, flown by a Cuban defector, landed there unchallenged, while Air Force One was parked on a nearby runway. Nineteen members of an Army detachment were arrested on pot charges at a Nike Hercules base on Mount Gleason, overlooking Los Angeles. One of them had been caught drying a large amount of marijuana on land belonging to the U.S. Forest Service. Three enlisted men at a Nike Hercules base in San Rafael, California, were removed from guard duty for psychiatric reasons. One of them had been charged with pointing a loaded rifle at the head of a sergeant. Although illegal drugs were not involved in the case, the three men were allowed to guard the missiles, despite a history of psychiatric problems. The squadron was understaffed, and its commander feared that hippies — “people from the Haight-Ashbury” — were trying to steal nuclear weapons.

More than one fourth of the crew on the USS Nathan Hale, a Polaris submarine with sixteen ballistic missiles, were investigated for illegal drug use. Eighteen of the thirty-eight seamen were cleared; the rest were discharged or removed from submarine duty. A former crew member of the Nathan Hale told a reporter that hashish was often smoked when the sub was at sea. The Polaris base at Holy Loch, Scotland, helped turn the Cowal Peninsula into a center for drug dealing in Great Britain. Nine crew members of the USS Casimir Pulaski, a Polaris submarine, were convicted for smoking marijuana at sea. One of the submarine tenders that docked at the base, the USS Canopus, often carried nuclear warheads and ballistic missiles. The widespread marijuana use among its crew earned the ship a local nickname: the USS Cannabis.

Four SAC pilots stationed at Castle Air Force Base near Merced, California, were arrested with marijuana and LSD. The police who raided their house, located off the base, said that it resembled “a hippie type pad with a picture of Ho Chi Minh on the wall.” At Seymour Johnson Air Force Base in Goldsboro, North Carolina, 151 of the 225 security police officers were busted on marijuana charges. The Air Force Office of Special Investigations arrested many of them leaving the base’s nuclear weapon storage area. Marijuana was discovered in one of the underground control centers of a Minuteman missile squadron at Malmstrom Air Force Base near Great Falls, Montana. It was also found in the control center of a Titan II launch complex about forty miles southeast of Tucson, Arizona. The launch crew and security officers at the site were suspended while investigators tried to determine who was responsible for the “two marijuana cigarettes.”

The true extent of drug use among American military personnel with access to nuclear weapons was hard to determine. Of the roughly 114,000 people who’d been cleared to work with nuclear weapons in 1980, only 1.5 percent lost that clearance because of drug abuse. But the Personnel Reliability Program’s 98.5 percent success rate still allowed at least 1,728 “unreliable” drug uses near the weapons. And those were just the ones who got caught.

Before assuming command of the 308th Strategic Missile Wing at Little Rock Air Force Base, Colonel John Moser had supervised a major drug bust at Whiteman Air Force Base, near Knob Noster, Missouri. More than 230 airmen were arrested for using and selling drugs there. Many were responsible for guarding and maintaining nuclear weapons. Some admitted to using marijuana, cocaine, and LSD on the job. Two of the three officers who were arrested had highly sensitive jobs at the base: they entered target information into the guidance systems of Minuteman missiles. When Moser arrived at Little Rock to assume command of the 308th, another drug bust was unfolding. Marijuana had been found in the control center at a Titan II complex. But the arrests didn’t end the drug use. The Strategic Air Command wasn’t immune to larger social forces, in an era before mandatory urine tests. Although launch officers rarely condoned illegal drug use, they spent alerts underground, without video cameras to reveal what was happening throughout a launch site. Their ability to command and control had its limits. Every so often, PTS crews would sit outside at a Titan II complex, light up a joint, crack open a few beers, and unwind at the end of a long day.

HENRY KISSINGER HAD TRIED to get rid of the Titan II. He considered the missile “inaccurate and unreliable.” It was a weapon system, he later explained, “which the Pentagon had been wanting to scrap for years and I had kept in service for trading purposes.” In 1972, while serving as the national security adviser to President Richard M. Nixon, Kissinger had offered a deal to the Soviet Union: the United States would decommission its Titan II missiles, if the Soviets agreed to retire their SS-9 missiles. The deal would eliminate a powerful threat to Moscow. And the Soviet missile was similar in a number of respects to the Titan II, employing the same type of fuel and oxidizer. But the SS-9 was also newer, larger, and capable of delivering a much heavier payload. The Soviet Union declined the offer. The Nixon administration was stuck with the Titan II — getting rid of fifty-four ballistic missiles, without getting anything in return from the Soviets, made little sense in the midst of an arms race.

The failed attempt to decommission an aging weapon system reflected the new balance of power. Robert McNamara had assumed that once the Soviet Union felt confident about its ability to destroy the United States in any nuclear exchange, it would stop building new missiles. But the Soviets didn’t share McNamara’s faith in mutually assured destruction. After the humiliation of the Cuban Missile Crisis, one of their diplomats had told an American counterpart, “You Americans will never be able to do this to us again.” In a rivalry where a nation’s power was measured numerically in warheads and bombs, the Soviet Union now sought to gain the upper hand. Within a decade of removing strategic weapons from Cuba, the Soviet Union increased the number of its long-range, land-based missiles from about 56 to more than 1,500. Its arsenal of submarine-based missiles rose from about 72 to almost 500. By the early 1970s, the Soviets had more long-range missiles than the United States. An elaborate antiballistic missile system had been created to defend Moscow. And a network of underground bunkers had been constructed beneath the city to protect the leadership of the Communist Party. Linked by secret subway lines, the bunkers could house thousands of people.

Although the United States possessed fewer ballistic missiles than the Soviet Union, it still had more nuclear weapons. McNamara had imposed a limit on the number of missiles that the United States would deploy — but not on the number of warheads that each missile could carry. Before leaving office, he’d approved the development of “multiple independently targetable reentry vehicles” (MIRVs). Publicly justified as a method of overwhelming a Soviet antiballistic missile system — adding more warheads to a single missile was less expensive than building more missiles — MIRVs also increased the number of Soviet targets that the United States could destroy in a first strike.

The Minuteman III missile, introduced in 1970, carried three warheads. They were housed on a post-boost vehicle, nicknamed the “bus,” that had its own rockets and guidance system. The bus separated from the missile and released each warhead over a different target, delivering them one after another, like a school bus dropping off children after school. The Poseidon missile, first deployed on American submarines in 1971, could carry fourteen warheads.

Kissinger was considered one of America’s leading authorities on nuclear strategy. For more than a decade his writing had helped to shape the national debate on the subject. He had served as an adviser to the Kennedy administration during the Berlin crisis. He knew as much as any civilian about the competing theories of nuclear warfare. And yet Kissinger was astonished by his first formal briefing on the SIOP. The smallest attack option would hit the Soviet Union with almost two thousand weapons; the largest with more than three thousand. The vast scale and inflexibility of the SIOP led Kissinger to describe it as a “horror strategy.” At a national security meeting in the Situation Room of the White House, he later wondered “how one rationally could make a decision to kill 80 million people.” President Nixon was equally appalled.

Most of the targets in the SIOP were still part of the Soviet war machinery — missile sites, air bases, command centers, ports. But the desire for assured destruction of the Soviet economy inspired calculations that made Fred Iklé’s theories about urban bombing seem like a relic of the Stone Age. RAND had developed a computer model to provide speedy estimates of the casualties and deaths that would be caused by different nuclear attacks. It was called QUICK COUNT. The types of weapons to be used in an attack, their targets, the prevailing winds, and the density of the local population were entered into an IBM-7090 computer — and then QUICK COUNT produced graphs, charts, and summaries of the potential carnage. It predicted the consequences of various attacks not only on the Soviet Union, but also on Eastern Europe, Western Europe, and the United States. And it included, as a bonus, an “Urban DGZ Selector” that helped war planners maximize the destruction of cities, allowing them to select the desired ground zeros likely to kill the most people.

A government report later outlined the “obstacle course to recovery” that victims of such nuclear attacks would have to navigate:

Although the human toll would be grim, the authors of the report were optimistic about the impact of nuclear detonations on the environment. “No weight of nuclear attack which is at all probable could induce gross changes in the balance of nature that approach in type or degree the ones that human civilization has already inflicted on the environment,” it said. “These include cutting most of the original forests, tilling the prairies, irrigating the deserts, damming and polluting the streams, eliminating certain species and introducing others, overgrazing hillsides, flooding valleys, and even preventing forest fires.” The implication was that nature might find nuclear warfare a relief.

Kissinger had once thought that Western Europe could be defended with tactical nuclear weapons, confining the damage to military targets and avoiding civilian casualties. But that idea now seemed inconceivable, and the refusal of America’s NATO allies to build up their conventional forces ensured that a military conflict with the Soviet Union would quickly escalate beyond control. During a meeting in the White House Situation Room, Kissinger complained that NATO nuclear policy “insists on our destruction before the Europeans will agree to defend themselves.”

Nixon’s administration soon found itself in much the same position as Kennedy’s, urgently seeking alternatives to an all-out nuclear war with the Soviet Union. “I must not be — and my successors must not be — limited to the indiscriminate mass destruction of enemy civilians as the sole possible response,” President Nixon told Congress. Phrases like “flexible response” and “graduated escalation” and “pauses for negotiation” seemed relevant once again, as Kissinger asked the Joint Chiefs of Staff to develop plans for limited nuclear war. But the Joint Chiefs still balked at making changes to the SIOP — and resisted any civilian involvement in target selection. The debacle in Vietnam had strengthened their belief that once the United States entered a war, the military should determine how to fight it. When Kissinger visited the headquarters of the Strategic Air Command to discuss nuclear war plans, General Bruce K. Holloway, the head of SAC, deliberately hid “certain aspects of the SIOP” from him. The details about specific targets were considered too important and too secret for Kissinger to know.

The Pentagon’s reluctance to allow civilian control of the SIOP was prompted mainly by operational concerns. A limited attack on the Soviet Union might impede the full execution of the SIOP — and provoke an immediate, all-out retaliation by the Soviets. A desire to fight humanely could bring annihilation and defeat. More important, the United States still didn’t have the technological or administrative means to wage a limited nuclear war. A 1968 report by the Weapons Systems Evaluation Group said that within five to six minutes of launching a submarine-based missile, the Soviet Union could “with a high degree of confidence” kill the president of the United States, the vice president, and the next fourteen successors to the Oval Office. The World Wide Military Command and Control System had grown to encompass eight warning systems, sixty communications networks, one hundred command centers, and 70,000 personnel. But the ground stations for its early-warning satellites could easily be destroyed by conventional weapons or sabotage, eliminating the ability to detect Soviet missile launches.

The National Emergency Airborne Command Post — a converted Boeing 747, designed to take off, whisk the president away from Washington safely, and permit the management of nuclear warfare in real time — did not have a computer. The officers manning the plane would have to record information about a Soviet attack by hand. And the entire command-and-control system could be shut down by the electromagnetic pulse and the transient radiation effects of a nuclear detonation above the United States. Communications might be impossible for days after a Soviet attack.

The system had already proven unreliable in conditions far less demanding than a nuclear war. In 1967, during the Six Day War, urgent messages warning the USS Liberty to remain at least one hundred miles off the coast of Israel were mistakenly routed to American bases in the Philippines, Morocco, and Maryland. The spy ship was attacked by Israeli planes almost two days after the first urgent warning was sent — and never received. The following year, when the USS Pueblo was attacked by North Korean forces, its emergency message calling for help took more than two hours to pass through the WWMCCS bureaucracy and reach the Pentagon. The American naval commander in Japan who managed to contact the Pueblo couldn’t establish direct communications with the Pentagon, the Situation Room at the White House, or commanders in the Pacific whose aircraft might have defended the ship.

During a conflict with the Soviet Union, messages would have to be accurately relayed within moments of an attack. A decade after the Kennedy administration recognized the problem, despite the many billions of dollars that had been spent to fix it, the command-and-control system of the United States was still incapable of managing a nuclear war. “A more accurate appraisal,” a top secret WSEG study concluded in 1971, “would seem to be that our warning assessment, attack assessment, and damage assessment capabilities are so limited that the President may well have to make SIOP execution decisions virtually in the blind, at least so far as real time information is concerned.” A few years later another top secret report said that the American response to a nuclear attack would be imperfect, poorly coordinated, and largely uncontrolled, with “confused and frightened men making decisions where their authority to do so was questionable and the consequences staggeringly large.”

AS THE SOVIET UNION ADDED multiple warheads to its ballistic missiles, Pentagon officials began to worry about the vulnerability of America’s nuclear forces. A Soviet surprise attack might wipe out not only the nation’s command-and-control facilities but also its land-based missiles. To deter such an attack, the Strategic Air Command considered a new retaliatory option, known as “launch on warning” or “launch under attack.” As soon as a Soviet attack was detected — and before a single warhead detonated — the United States would launch its land-based missiles, saving them from destruction. A launch-on-warning policy might dissuade the Kremlin from attempting a surprise attack. But it would also place enormous demands on America’s command-and-control system.

Missiles launched from Soviet submarines could hit Minuteman and Titan II bases in the central United States within about fifteen minutes; missiles launched from the Soviet Union would arrive in about half an hour. The president would have no more than twenty minutes to decide whether to retaliate — and would probably have a lot less time than that. With each passing minute, the pressure to “use it or lose it” would grow stronger. And the time constraints would increase the risk of errors. The reliability of America’s early-warning system attained an existential importance. If the sensors failed to detect a Soviet attack, the order to launch might never be given. But if they issued an attack warning erroneously, millions of people would be killed by mistake.

The Pentagon decision to provide the United States with a nuclear hair trigger, capable of being fired at a moment’s notice, oddly coincided with the warmest relations between the two superpowers since the end of the Second World War. The Soviet invasion of Czechoslovakia in 1968 hadn’t raised tensions between the two Germanys, inspired massive demonstrations against the Soviet Union, or provoked much European revulsion toward communism. On the contrary, the overthrow of a moderate Czech government had encouraged Willy Brandt, the foreign minister of West Germany at the time, to seek closer ties with the Soviet Union. The status quo in Europe, the division between East and West, would not be challenged.

Within a few years, a series of international agreements clarified the legal status of Berlin, recognized the sovereignty of both German governments, promised to reduce the threat of nuclear war, and established a working relationship between the United States and the Soviet Union known as détente. The two countries signed the AntiBallistic Missile Treaty, allowing each side to defend two locations from attack; the Threshold Test Ban Treaty, limiting the size of underground detonations to 150 kilotons; and an Interim Agreement on Certain Measures with Respect to the Limitation of Strategic Offensive Arms, freezing the number of land-based ballistic missiles and permitting the deployment of new submarine-based missiles only when old ones were retired.

The advent of détente did not, however, end the nuclear arms race. The United States and the Soviet Union continued to modernize weapon systems and improve their accuracy. More than ever, nuclear weapons seemed important as totems of status and world power. Not long after taking office, President Nixon tried to end the Vietnam War by threatening the use of nuclear weapons, convinced that Eisenhower had employed a similar tactic to end the war in Korea. “I call it the Madman Theory, Bob,” Nixon told his chief of staff, H. R. Haldeman. “I want the North Vietnamese to believe that I’ve reached the point where I might do anything to stop the war.” The secretary of state, the secretary of defense, and the Joint Chiefs of Staff thought it was a bad idea. But Nixon and Kissinger thought the plan might work. Ignoring the safety risks, the Strategic Air Command secretly resumed its airborne alert for two weeks. B-52s loaded with hydrogen bombs took off from bases in the United States and flew circular routes along the coast of the Soviet Union. Neither the Soviets nor the Vietcong was fooled by the bluff.

A few years later, at the height of the 1973 Arab-Israeli War, nuclear weapons were once again utilized as a diplomatic tool. Concerned that the Soviet Union might send troops to Egypt, Secretary of State Kissinger and Secretary of Defense James R. Schlesinger placed American military forces throughout the world at DEFCON 3. The elevated level of readiness was a signal to the Soviet Union, implying that the United States was willing to fight a nuclear war over the issue. The Soviets didn’t intervene in the Mideast conflict, and Kissinger later attributed their reluctance to the administration’s bold diplomacy. Great leaders sometimes need to appear unbalanced, he thought: “What seems ‘balanced’ and ‘safe’ in a crisis is often the most risky.”

Fred Iklé served as the head of the U.S. Arms Control and Disarmament Agency during the Nixon and Ford administrations. Iklé brought to the job an extensive knowledge of nuclear weapons, deterrence theory, and the workings of the command-and-control system. He argued against the adoption of a launch-on-warning policy, worried that it could inadvertently prove to be disastrous. Nevertheless, the policy had a strong military and psychological appeal. “Launching the ICBM force on attack assessment is probably the simplest and most cost-effective way to frustrate a [Soviet] counterforce attack,” a classified RAND report noted. “But as a declared policy, we believe it would be vigorously opposed as both dangerous and unstable (an accident could theoretically precipitate a nuclear war).”

At a meeting of the National Security Council, Iklé expressed his opposition to launch on warning, calling it “accident-prone.” Secretary of State Kissinger disagreed, praising its usefulness as a deterrent. Kissinger felt confident that the command-and-control system could handle it and stressed that “the Soviets must never be able to calculate that you plan to rule out such an attack.” The national security adviser, Brent Scowcroft, agreed with Kissinger. Reason now played a diminished role in nuclear strategy. “It is not to our disadvantage,” Scowcroft said, “if we appear irrational to the Soviets in this regard.”

Too much madness, however, could be dangerous. Since the days of Harry Truman, the president of the United States had been entrusted with the sole authority to order the use of nuclear weapons. It gave one human being the ability to destroy cities, nations, entire civilizations. The president was accompanied everywhere by a military aide carrying the “football” — a briefcase that held the SIOP Decisions Handbook, a list of secret command bunkers throughout the United States, and instructions on how to operate the Emergency Broadcast System. The SIOP Decisions Handbook outlined various attack options, using cartoonlike illustrations to convey the details quickly. It was known as the Black Book.

Eager to defend the civilian control of nuclear weapons from military encroachment, John F. Kennedy and Robert McNamara had fought hard to ensure that only the president could make the ultimate decision. But they hadn’t considered the possibility that the president might be clinically depressed, emotionally unstable, and drinking heavily — like Richard Nixon, during his final weeks in office. Amid the deepening Watergate scandal, Secretary of Defense Schlesinger told the head of the Joint Chiefs to seek his approval before acting on “any emergency order coming from the president.” Although Schlesinger’s order raised questions about who was actually in command, it seemed like a good idea at the time.

The Wrong Tape

One month after the inauguration of President Jimmy Carter, a member of his national security staff, General William E. Odom, attended briefings on the SIOP at the headquarters of the Strategic Air Command in Omaha. Odom was considered a staunch anti-Communist, one of the hard-liners in the new administration. He was a Soviet expert, fluent in Russian, who’d attended West Point and trained as a tactical nuclear targeting officer for the Army. His visit to SAC headquarters occurred in February 1977. Eight years had passed since Henry Kissinger began to push for more flexibility in the SIOP. Secretary of Defense Schlesinger had announced in 1974 that America’s war plans were being revised, that they would soon include “Limited Nuclear Options” and “Regional Nuclear Options” using fewer weapons. And yet General Odom could find no trace of those changes in the SIOP. Like others before him, nuclear initiates granted a secret knowledge, Odom was stunned by the SIOP:

A policy of launch on warning was “absurd and irresponsible,” and implementing the SIOP under any conditions would be “the height of folly.” The SIOP now called for the Soviet Union to be hit with about ten thousand nuclear weapons. But what disturbed Odom the most about the Joint Strategic Target Planning Staff in Omaha was that they didn’t seem to have any postattack plans: “Things would just cease in their world about 6 to 10 hours after they received the order to execute the SIOP.”

President Carter was determined to end the arms race with the Soviet Union. And he knew more about nuclear weapons than any of his predecessors at the White House, except, perhaps, Eisenhower. Carter had attended the U.S. Naval Academy, served as an officer on submarines, and helped to design the first nuclear propulsion systems for the Navy. A few weeks before his inauguration, Carter had met with the Joint Chiefs of Staff and asked them an unexpected question: How long would it take to reduce America’s nuclear arsenal to just one or two hundred ballistic missiles? The room fell silent — and no answer was given.

In that moment, President Carter had revealed himself to be an advocate of “minimum deterrence,” a strategy that the Navy had endorsed in the late 1950s, as the Polaris submarine was being developed. He thought that one or two hundred missiles might be sufficient to deter the Soviets. And if both superpowers reduced their strategic forces to those levels, neither could launch a successful first strike. During his inaugural address, Carter spoke about his ultimate goal: “the elimination of all nuclear weapons from this Earth.” To make sure the issue was never far from his mind, he kept wooden miniatures of Soviet and American missiles on his desk in the Oval Office.

The Joint Chiefs of Staff regarded Carter with suspicion. The new president not only supported minimum deterrence, he also sought a ban on all nuclear testing. He proposed large cuts in military spending. He sincerely wanted new arms control agreements, world peace, friendship with the Soviet Union. And he appointed Harold Brown — one of McNamara’s former whiz kids — to serve as secretary of defense. Brown thought that the United States hadn’t fallen behind the Soviets and that new strategic weapons, like the B-1 bomber, weren’t urgently needed. Within weeks of taking office, Carter found his plans opposed by most Republicans, many Democrats, the armed services — and even the Soviets. At the Kremlin, his proposal to accelerate the reduction of ballistic missiles seemed like an attempt to gain favorable publicity, and his criticism of human rights violations in the Soviet Union were regarded as insulting. The Soviet leadership much preferred dealing with Nixon and Kissinger, who never mentioned the repression of dissidents.

A new organization, the Committee on the Present Danger, soon attacked the Carter administration for being weak on defense and endangering the security of the United States. The group’s membership included academics, defense intellectuals, former government officials, and retired military officers. They warned that within a few years the nation would face a “window of vulnerability,” a period in which the Soviets might be able to launch a surprise attack that spared American cities but destroyed all of its land-based missiles. The president would then face an agonizing choice: accede to the demands of the Soviet Union and save American lives — or launch submarine-based missiles at Soviet cities and cause pointless, mutual annihilation. The committee’s views were succinctly expressed in an essay by Richard Pipes, a history professor at Harvard and one of the group’s founders: “Why the Soviet Union Thinks It Could Fight and Win a Nuclear War.” The Soviets were violent, deceitful, authoritarian, and cunning, Pipes argued, and they’d already shown a willingness to commit mass murder on behalf of communism. The downfall of the United States now seemed within their grasp and would be pursued, regardless of the cost.

The window of vulnerability — like the bomber gap and the missile gap before it — provided a strong rationale for increased spending on defense. And like those other scares, it was based more on fear than on facts. A successful surprise attack on America’s land-based missiles wouldn’t be easy to pull off. To achieve a 95 percent certainty of wiping them out, at least two Soviet warheads would have to be aimed at each silo. Those warheads would have to land in precisely timed intervals, so that the blast effects of one didn’t destroy the other. And the Soviets would have to prevent the Strategic Air Command from launching its missiles on warning. Even if the surprise attack were successful, disabling every single Minuteman and Titan II, the fallout from the nuclear blasts would kill somewhere between two million and twenty million Americans. And the United States would still have thousands of nuclear warheads, mounted on submarine-based missiles, ready to seek revenge.

President Carter’s idealistic vision soon collided with the reality of the late 1970s. He had to contend with gasoline shortages, high unemployment, and inflation; anxieties about the decline of American power; the arms buildup in the Soviet Union, its crackdown on dissidents, its use of Cuban troops as proxies in Ethiopia and Angola. The Senate refused to approve another arms control treaty, and détente became a thing of the past. Instead of cutting the defense budget, Carter increased it for the first time in more than a decade. Instead of adopting a strategy of minimum deterrence, he endorsed a “countervailing strategy” that would allow the president to use limited nuclear strikes in a variety of situations. Instead of eliminating strategic weapons, he backed the development of entirely new ones — the MX long-range missile, the Pershing II medium-range missile, cruise missiles that used jet engines instead of rockets to fly low and evade Soviet radar, the B-2 bomber, the Trident submarine.

The MX missile system embodied the strategic thinking of its time. To avoid destruction in a surprise attack, the MX would be mounted on a two-hundred-foot-long truck. The missile would constantly be moved between twenty-three protective concrete shelters, like a pea in an immense shell game. The Soviet Union would never know which shelter housed a missile. The shelters would be a mile apart. Twenty-two of them would contain fake missiles — and those decoys would also be moved constantly by truck. If the scheme worked, the Soviets would have to use at least forty-six warheads to destroy a single MX missile.

President Carter approved the deployment of two hundred MX missiles in the Great Basin area of Utah and Nevada. The missiles would be scattered across roughly fifteen thousand square miles of federal land, most of it closed to the public. Eight thousand miles of new roads would be built for access to the MX sites. About a hundred thousand workers would be required to construct the system and about half that number to run it. The total cost of the project was estimated to be at least $40 billion. The new weapon was designed not only to close the window of vulnerability for the United States but also to open one for the Soviet Union. Each MX would carry ten highly accurate warheads, thereby placing Soviet missiles at risk of destruction during an American first strike.

AT ABOUT ELEVEN O’CLOCK in the morning on November 9, 1979, the computers at the NORAD headquarters inside Cheyenne Mountain said that the United States was under attack. The huge screen in the underground command center at SAC headquarters showed that Soviet missiles had been launched from submarines off the West Coast. The same message was received by computers in the National Military Command Center at the Pentagon and the Alternate National Military Command Center at Site R inside Raven Rock Mountain. And then more missiles appeared on the screen, launched not only from submarines but also from sites within the Soviet Union. It was a massive attack, and warheads would begin to hit American targets within five or six minutes.

Whenever NORAD’s early-warning sensors detected signs of a possible missile launch, a Missile Display Conference was held. It happened about four times a day; the infrared sensors on the Air Force satellites could be triggered by forest fires, volcanic eruptions, and other sources of heat. The officers on duty would discuss whether the threat seemed real or merely a false alarm. The commander in chief of NORAD would decide if a Threat Assessment Conference had to be arranged, bringing the head of SAC and the chairman of the Joint Chiefs of Staff into the discussion. That type of conference happened once or twice a week. And if missiles truly seemed to be heading toward the United States, a Missile Attack Conference would be set up. It would give the president a chance to speak with senior officers, listen to their advice, and decide whether to launch missiles in retaliation. A Missile Attack Conference had never been held.

As the computer screens at NORAD filled with Soviet missiles, a Threat Assessment Conference was called. Although the pattern of the attack seemed to fit with the Pentagon’s assumptions about Soviet war plans, its timing made little sense. Tensions between the superpowers weren’t particularly high, and nothing in the news seemed to warrant a “bolt from the blue” attack on the United States. Duty officers at NORAD contacted the radar and ground stations whose sensors were relaying information about the launches. None of them had detected signs of any missiles. The NORAD computers seemed to be providing an erroneous — but highly realistic — account of a Soviet surprise attack.

As a precaution, the Klaxons were sounded at SAC bases nationwide. Bomber crews ran to their planes, and missile crews were put on heightened alert. Fighter-interceptors took off to look for signs of a Soviet attack. The National Emergency Airborne Command Post left Andrews Air Force Base — without President Carter on board. And air traffic controllers throughout the country prepared to clear America’s airspace for military flights, warning every commercial airliner that it might soon have to land.

As the minutes passed without the arrival of Soviet warheads, it became clear that the United States wasn’t under attack. The cause of the false alarm was soon discovered. A technician had put the wrong tape into one of NORAD’s computers. The tape was part of a training exercise — a war game that simulated a Soviet attack on the United States. The computer had transmitted realistic details of the war game to SAC headquarters, the Pentagon, and Site R.

The computers at NORAD had been causing problems for more than a decade. Although they were perhaps the most important data-processing machines in the United States — responsible for compiling and assessing information from all its early-warning radars and satellites — the Honeywell 6060 computers were already obsolete when NORAD installed them within Cheyenne Mountain. A 1978 investigation by the General Accounting Office (GAO) found that budget cuts and bureaucratic inflexibility during the Nixon administration had forced NORAD to buy the computers — despite protests from the head of NORAD that they lacked sufficient processing power for crucial early-warning tasks. NORAD’s computers were frequently out of commission, the GAO reported, “due to the lack of readily available spare parts.” Many of the parts hadn’t been manufactured by Honeywell for years.

The morale at NORAD, like its aging computers and software, left room for improvement. A couple of months after the false alarm, twenty-three security officers assigned to the Combat Operations Center inside Cheyenne Mountain were stripped of their security clearances. According to the Air Force Office of Special Investigations, the security force responsible for protecting the nerve center of America’s command-and-control system was using LSD, marijuana, cocaine, and amphetamines.

“FALSE ALARM ON ATTACK SENDS FIGHTERS INTO SKY” was one of the headlines, when news of the training tape incident leaked. Pentagon officials denied that the missile warning had been taken seriously. But the technical and human errors at NORAD felt in keeping with the general mood of the country. An accidental nuclear war didn’t sound inconceivable to most people — America seemed to be falling apart. A few months earlier a nuclear reactor at Three Mile Island in Pennsylvania had suffered a partial meltdown, largely because a worker at the plant had turned off an emergency cooling system by mistake.

At about two thirty in the morning on June 3, 1980, Zbigniew Brzezinski, the president’s national security adviser, was awakened by a phone call from a staff member, General William E. Odom. Soviet submarines have launched 220 missiles at the United States, Odom said. This time a surprise attack wasn’t implausible. The Soviet Union had recently invaded Afghanistan, confirming every brutal stereotype promoted by the Committee on the Present Danger. The United States was leading a boycott of the upcoming Moscow Olympics, and relations between the two superpowers were at their lowest point since the Cuban Missile Crisis. Brzezinski told Odom to call him back with confirmation of the Soviet attack and its intended targets. The United States would have to retaliate immediately; once the details of the attack were clear, Brzezinski would notify the president. Odom called back and said that 2,200 missiles were heading toward the United States — almost every long-range missile in the Soviet arsenal. As Brzezinski prepared to phone the White House, Odom called again. The computers at NORAD said that Soviet missiles had been launched, but the early-warning radars and satellites hadn’t detected any. It was a false alarm. Brzezinski had allowed his wife to sleep through the whole episode, preferring that she not be awake when the warheads struck Washington.

SAC bomber crews had run to their planes and started the engines. Missile crews had been told to open their safes. The airborne command post of the Pacific Command had taken off. And then the duty officer at the Pentagon’s National Military Command Center ended the Threat Assessment Conference, confident that no Soviet missiles had been launched. Once again, NORAD’s computers and its early-warning sensors were saying different things. The problem was clearly in one of the computers, but it would be hard to find. A few days later NORAD computers warned SAC headquarters and the Pentagon for a third time that the United States was being attacked. Klaxons sounded, bomber crews ran to their planes — and another Threat Assessment Conference declared another false alarm.

This time technicians found the problem: a defective computer chip in a communications device. NORAD had dedicated lines that connected the computers inside Cheyenne Mountain to their counterparts at SAC headquarters, the Pentagon, and Site R. Day and night, NORAD sent test messages to ensure that those lines were working. The test message was a warning of a missile attack — with zeros always inserted in the space showing the number of missiles that had been launched. The faulty computer chip had randomly put the number 2 in that space, suggesting that 2 missiles, 220 missiles, or 2,200 missiles had been launched. The defective chip was replaced, at a cost of forty-six cents. And a new test message was written for NORAD’s dedicated lines. It did not mention any missiles.

TOWARD THE END OF the Eisenhower administration, amid the fiery rhetoric of the missile gap, Bob Peurifoy became concerned that the Soviet Union might attack the United States. With help from his wife, Barbara, and a local contractor, Peurifoy built a bomb shelter underneath the garage at the family home in Albuquerque. Other engineers at Sandia added bomb shelters to their houses, too. The laboratory was a prime target for the Soviets, and the series of international crises during the first two years of the Kennedy administration made the decision seem wise. The Peurifoy shelter had food, water, a dosimeter to measure radiation levels, a door that could be sealed shut, a hand-cranked ventilation fan, a gun, and enough room for five people. He later viewed it as a youthful folly. When the family moved to another house in 1967, a few miles from the nuclear weapon storage facility at Site Able, he didn’t bother to build another shelter. Peurifoy couldn’t dig a hole deep enough to protect his family from the thermonuclear warheads likely to hit the neighborhood. And by the mid-1970s, he was preoccupied with a different threat. Although Peurifoy was conservative and anti-Communist, a Republican and a supporter of increased spending on defense, the nuclear weapons in the American arsenal were the ones keeping him up at night.

The Fowler Letter’s only immediate effect was to raise the possibility that Glenn Fowler would lose his job. His urgent safety warning didn’t persuade the Air Force to remove nuclear weapons from its bombers on ground alert. At the Department of Defense and the Atomic Energy Commission, the anger provoked by the letter was intense. High-ranking officials from both organizations flew from Washington, D.C., to meet with the head of Sandia. In preparation for the meeting, Peurifoy asked Stan Spray to put together an exhibit of weapon components that had been subjected to abnormal environments. Perhaps seeing would be believing: the melted solder on charred circuit boards seemed like irrefutable evidence that nuclear weapons could behave unpredictably during a fire. Spray’s presentation was soon known as the Burned Board briefing. Donald R. Cotter, the assistant to the secretary of defense for atomic energy, and Major General Ernest Graves, the AEC official to whom Fowler’s letter had been sent, weren’t impressed. They found the evidence unconvincing. And they were outraged that Sandia had put these claims on the record. The American stockpile contained dozens of different types of nuclear weapons, and the Fowler Letter didn’t assert there was a minor safety problem with one of them. It suggested that none were demonstrably safe.

Don Cotter was particularly upset. He knew Peurifoy and Bill Stevens well. Before going to the Pentagon, Cotter had worked at Sandia for years. He’d designed the electrical systems of nuclear weapons, championed early safety devices, and helped Fred Iklé prepare the RAND report on weapon safety. Cotter was offended by the Fowler Letter. His response to it was blunt: “It’s our stockpile. We think it’s safe. Who do you guys think you are?” Peurifoy’s team had challenged not only the conventional wisdom about weapon design but also the readiness of some NATO units and the Strategic Air Command.

Fowler kept his job. But the recommendations in his letter weren’t followed. No air-delivered weapons were taken out of service or retrofitted with new safety mechanisms. Instead, a series of government studies was commissioned to explore the issue of nuclear weapon safety, a classic bureaucratic maneuver to delay taking any action. The Department of Defense argued that “the magnitude of the safety problems is not readily apparent” — and it now had unprecedented influence over the nuclear stockpile. The Atomic Energy Commission was disbanded in 1975. It was replaced by the Energy Research and Development Administration, an agency that lasted only two years, before being subsumed into the Department of Energy. The Joint Committee on Atomic Energy — which had served for three decades as a powerful civilian counterweight to the military — was abolished in 1977. The Pentagon wielded largely unchecked power over the management of nuclear weapons, and its Defense Nuclear Agency had a set of priorities that differed from Bob Peurifoy’s. “The safety advantages gained by retrofitting existing stockpile weapons… will be a costly program that in all probability will reduce funds available for future weapons,” the DNA said.

The Air Force deployed most of the weapons that Peurifoy wanted to fix. And it supported the use of new safety devices, so long as they didn’t require:

1. Modification of any current operational aircraft

2. Additional crew actions and

3. Expenditure of Air Force money.

The Air Force also continued to have little interest in permissive action links or other forms of use control. The latest PALs were far more sophisticated and reliable than the ones provided to NATO in the early 1960s. The new Category D PALs had a six-digit code with a million possible combinations, a limited-try feature that permanently locked the weapon if the wrong numbers were entered, and the capability to store multiple codes. The president could now choose to unlock some nuclear weapons, but not others, by selecting a certain code. The system promised centralized, secure command and control. But the Strategic Air Command continued to resist installing PALs inside its warheads and bombs.

After the accident at Thule, the Pentagon had ordered SAC to impose some form of use control. Instead of relying on PALs, during the early 1970s the Air Force put a coded switch in the cockpit of every bomber that carried nuclear weapons. The switch permitted an arming signal to be sent to the bomb bay when the right code was entered. The lock had been placed on the bomber, not inside the bombs — and a stolen weapon could still be detonated with a simple DC signal. SAC was far more worried about its weapons being rendered inoperable during wartime than about someone stealing them or using them without proper authorization. During the late 1970s, a coded switch was finally placed in the control center of every SAC ballistic missile. It unlocked the missile, not the warhead. And as a final act of defiance, SAC demonstrated the importance of code management to the usefulness of any coded switch. The combination necessary to launch the missiles was the same at every Minuteman site: 00000000.

Peurifoy was undaunted by the many layers of bureaucratic opposition. The issue at stake wasn’t trivial, and he was determined to persuade others in the defense community that the danger was real. The cost of adding weak links and strong links and a unique signal mechanism was about $100,000 per weapon. The Office of Management and Budget estimated that the installation of those safety devices in the two most widely used Air Force bombs, the Mark 28 and the B-61, would cost about $360 million. Peurifoy realized that was a good deal of money — but a nuclear weapon accident could be a hell of a lot more expensive. The amount of money needed for that retrofit was roughly 1 percent of what the Air Force planned to spend driving around MX missiles in the Utah and Nevada desert. The Pentagon’s fixation on obtaining new weapons, instead of properly maintaining older ones, would be hard to overcome. But the fight seemed worthwhile. A friend sent Peurifoy a cartoon that showed a member of the Supreme Court speaking from the bench. It conveyed Peurifoy’s general attitude, when the facts were on his side. “My dissenting opinion will be brief,” the justice said. “You’re all full of crap.”

The role of the weapons laboratories had become mainly advisory. They competed for contracts from the Department of Defense — and felt reluctant to criticize their largest customer. Peurifoy had no authority to demand changes in weapon systems that the armed services already possessed. But he refused to sign the Sandia major assembly release of any new bombs or warheads that didn’t have the new safety devices. And without his approval, those weapons couldn’t enter the stockpile. In 1977, almost four years after gaining some real authority at the lab, Peurifoy signed the release papers on a modification of the B-61 bomb. It was the first nuclear weapon to feature weak link/strong link technology.

As the dispute with the Pentagon dragged on, Peurifoy learned that the armed services were no longer telling him about nuclear weapon accidents. Broken Arrows would be difficult to hide, but the more commonplace mishaps — short circuits, bombs falling off loading carts, weapon carriers overturned — weren’t being reported to him. Peurifoy would often hear about them through other sources. The sense of denial at the upper levels of the Air Force and the Department of Defense had a ripple effect throughout both institutions. The bomber crews, the missile crews, the technicians who routinely handled warheads and bombs, the maintenance teams and firefighters — they were told the weapons were perfectly safe. The misinformation placed them at greater risk. It was also a form of disrespect toward young servicemen and women who were already risking their lives. And it encouraged careless behavior around nuclear weapons. In many ways, denying the safety problems only made them worse.

While Peurifoy fought the bureaucratic wars, Bill Stevens and the rest of the nuclear safety department continued to study how to make nuclear weapons less likely to detonate by accident, spread plutonium, or fall into the wrong hands. During the late 1960s, Stevens had begun to worry about a terrorist attempt to steal a weapon, and the massacre at the 1972 Munich Olympics demonstrated that the threat was real. The weapons inside NATO storage igloos seemed the most vulnerable to theft, not only by potential terrorists but also by rogue elements of an allied army or enemy troops. If an igloo seemed on the verge of being overrun, NATO forces were supposed to “spike the guns” — to attach a shaped explosive charge to each weapon and blow it up. A nuclear detonation wouldn’t occur. But the collateral damage could be enormous, and a great deal of plutonium dust might be spread. Stevens thought that better ways of keeping weapons out of the wrong hands needed to be found and that the risk of plutonium dispersal had to be taken most seriously.

Changes were soon made to the storage practices at NATO igloos and to the emergency procedures for destroying weapons. Antiterrorism research at Sandia led to the development of new perimeter control technologies, such as motion detectors, and innovative methods for stopping intruders who somehow managed to get past the door of an igloo. Nozzles on the walls would rapidly fill the place with sticky foam, trapping intruders and preventing the removal of nuclear weapons. The foam looked ridiculous, like a prop from a Three Stooges film, but it worked.

Peurifoy and Stevens also looked at how nuclear weapons should be rendered safe after an accident. The civilians at Sandia and the military personnel in Explosive Ordnance Disposal units often had conflicting notions about what should be done. It was another dispute that pitted scientists in white lab coats against men in uniform. Air Force bomb squads were accustomed to dealing with conventional weapons. And they were trained to get the job done quickly — during wartime, an unexploded bomb near a runway could prevent essential aircraft from taking off. The EOD guys liked to approach a weapon, tear it down fast, and get rid of it. Peurifoy and Stevens thought that wasn’t a good idea with nuclear weapons. A hydrogen bomb that survived an accident reasonably intact could still detonate if someone handled it improperly. Even if it didn’t produce a nuclear yield, the high explosives could spread plutonium and harm anyone nearby.

After the B-52 crash near Cumberland, Maryland, an Air Force EOD team started to remove the weapons from the wreckage of the plane, using improvised heavy machinery — until a representative from Sandia intervened and asked them to stop. The bombs weren’t moved until their condition had been assessed. A naval bomb disposal team began to disassemble the Mark 28 bomb recovered from the ocean near Palomares — until another Sandia nuclear safety specialist made clear that a ship, rolling over swells, might not be the best place for the task. Peurifoy and Stevens thought that, most of the time, there was no need to rush. “Don’t move someone who’s hurt before you know the extent of the injuries,” a basic rule of first aid, also applied to nuclear weapons. Ease of disassembly had never been a top priority among weapon designers. In fact, it was rarely considered when weapons were on the drawing board. Inside the metal casing, parts were tightly welded or glued together. If you weren’t careful, thermal batteries could be ignited, high explosives set off. Peurifoy took an EOD course and gained tremendous respect for the soldiers and airmen who put on bomb suits to render bombs safe. They were fearless. But the weapons they typically handled might kill them and injure people within about a quarter of a mile. Peurifoy didn’t want anyone to feel hurried or gung ho while trying to dismantle a thermonuclear warhead.

The need to retrofit and retire older weapons in the stockpile became more urgent after a discovery about the Mark 28 hydrogen bomb. Stan Spray found that one of the bomb’s internal cables was located too close to its skin. If the weapon was exposed to prolonged heat, the insulation of the cable would degrade — and the wires inside it could short circuit. One of those wires was connected to the ready/safe switch, another to the thermal battery that charged the X-unit. It was a serious problem. The heat from a fire could arm a Mark 28 bomb, ignite its thermal battery, charge its X-unit, and then fully detonate the high explosives. Depending on the particular model of the Mark 28, a blast of anywhere from 70 kilotons to 1.5 megatons would immediately follow.

The problem with the Mark 28 was more significant than the safety flaws in other weapons. Mark 28 bombs were routinely carried by B-52 bombers on ground alert. And those B-52s sometimes caught on fire, even when they never left the ground. The bomber carried more than 300,000 pounds of highly flammable JP-4 jet fuel, a mix of gasoline and kerosene. In preparation for a typical B-52 flight, the crew would spend at least an hour in the plane, going through checklists, before starting the engines — and then the engines would be started one after another, until all eight were running. It could take an hour and a half for the pilot to get a B-52 into the air. But planes on ground alert were expected to be airborne within ten or fifteen minutes, the maximum time available for a “base escape.” Explosive cartridges on the four engine pods would be detonated by the copilot, as soon as he climbed into the plane, spinning the turbines rapidly and starting all eight engines in about a minute. A “cartridge start” was a memorable sight — a series of small explosions, B-52s filling the runway with clouds of smoke — and crews on ground alert practiced it regularly. And yet it could also start a fire.

The combination of Mark 28 bombs and B-52 bombers on alert was increasingly dangerous. Peurifoy doubted it was worth the risk. Both were aging weapon systems; many of the B-52s were older than their pilots. And most of the planes would probably never reach their targets, let alone return safely from a mission. After a 1975 briefing on the role of the Strategic Air Command’s bombers in executing the SIOP, the head of the CIA, William Colby, expressed surprise that “our B-52s are planned for one-way missions.” Once an emergency war order was transmitted, the bombers on ground alert would quickly take off from their bases in the United States, fly eight to ten hours toward Soviet targets — and find what? The Soviet Union would have already been hit by thousands of warheads delivered by American missiles. Targets that hadn’t been destroyed were likely to be surrounded by antiaircraft missiles, and dust clouds of unimaginable scale would blanket the landscape. Each B-52 was assigned a poststrike base in Europe or the Middle East where it was supposed to land, refuel, and pick up more nuclear weapons for another run at the Soviets. Would any of those bases still exist, if bombers somehow managed to survive their first passage through Soviet airspace? Most B-52 crews didn’t count on it.

Stan Spray added components from the Mark 28 bomb to his Burned Board briefing, along with a dramatic flourish: when the bomb’s wires short-circuited, a flashbulb went off. The briefing was given to hundreds of officials — with little immediate effect. A study of all the nuclear weapons in the American arsenal was completed by one of Peurifoy’s deputies in 1977. It provided the Department of Defense with a list of the weapons posing the greatest threat and a timetable for retiring them or improving their safety. The Mark 28 bomb was at the top of the list, followed by the W-25 warhead of the Genie antiaircraft missile. Despite being the oldest sealed-pit weapon in the stockpile, vulnerable to lightning, and fitted with an outdated accelerometer, the Genie was still being loaded onto fighter planes. On the list of weapons requiring urgent attention, the only strategic warhead was the W-53 atop the Titan II missile. It needed a “retrofit for Enhanced Electrical Safety.”

In 1979 the Department of Defense finally accepted some of the recommendations that Sandia’s safety department had been making for years — but didn’t want to pay for them. The Pentagon agreed to schedule retrofits of weapons like the Mark 28, so long as the cost wouldn’t interfere with the acquisition of new weapons. And until the funds were obtained, the Mark 28 could still be carried by B-52s on ground alert. Although the Air Force balked at devoting a few hundred million dollars to improve the safety of hydrogen bombs, it planned to spend at least $10 billion to equip B-52s with cruise missiles. Instead of trying to penetrate Soviet airspace, the bombers would launch cruise missiles a thousand miles from their targets, turn around, and come home. Until those cruise missiles were available, B-52s were loaded with Short-Range Attack Missiles (SRAMs), carried in a rotary rack. It turned as each missile was fired, like the cylinder of a revolver shooting bullets. The SRAMs were designed to fly a hundred miles or so, destroy Soviet air defenses, and give the B-52 a better chance of reaching its target. The missiles had a destructive force of as much as 200 kilotons, and a single B-52 could carry a dozen of them.

Peurifoy was frustrated by the delays. Even the retrofit of the Mark 28, top on the list, kept getting pushed back. Through a friend in the Air Force, Peurifoy arranged for General Howard W. Leaf to visit Sandia on June 13, 1980. Leaf would be given the Burned Board briefing. The safety problems with the Mark 28 would be outlined in detail, as well as the history of nuclear weapon accidents and the development of weak link/strong link devices. Leaf had an important job, inspector general of the Air Force, with the authority to cut through red tape. The false alarm caused by a faulty computer chip at NORAD, ten days earlier, had brought renewed attention to the importance of command and control, the limits of technology, the risks of human error. After lengthy meetings at Sandia, General Leaf returned to Washington, D.C. — and commissioned another study on the safety of the Mark 28 bomb.

ON SEPTEMBER 15, 1980, Jeffrey A. Zink was pulling an alert at Grand Forks Air Force Base in North Dakota. Zink was the navigator of a B-52. Once a month he and the rest of his crew would sleep in a building at the end of a runway, with a tunnel leading to their plane. Four or five other B-52 crews would stay there, too, along with the crews of their tankers. In some respects it felt like being confined in a prison. The alert quarters were surrounded by concertina wire, motion detectors, and security police carrying M-16s. Zink and his friends spent most of their time being bored. They would eat, sleep, read books, take naps, watch crap like The Love Boat on TV. But Zink always thought boredom, in this case, was good. Boredom meant that deterrence still worked. So long as these fifty young men were stuck there doing nothing, America’s nuclear strategy was a success. About once a week, however, the Klaxons would sound, and life would suddenly become more interesting.

Zink had never intended to join the Air Force. In the mid-1970s he was a longhaired, true-blue hippie attending the University of Pittsburgh and planning to go to law school. One day he walked into an Air Force recruiter’s office, thinking it would be cool to fly planes. The recruiter told Zink that his eyes weren’t good enough to become a pilot — but he could become a navigator. Zink put aside law school and joined the Air Force in 1977, right after graduation. His hippie girlfriend was stunned, and their relationship soon ended. At first, Zink didn’t fit into the tough, regimented culture of the Strategic Air Command. “What have I gotten myself into?” he wondered. “I don’t think I like these people.” But his feelings gradually changed, and he eventually became a lieutenant colonel.

The navigator of a B-52 sat at a desktop in the “chin” of the plane, a lower level beneath the pilot. Beside the navigator sat the bombardier. They both had ejection seats that fired downward. Their compartment was small, cramped, and windowless, with a ceiling about five feet high. Training flights lasted six to eleven hours, and they could be rough. The eight engines were so loud that the navigator and the bombardier, seated a foot or two from each other, couldn’t shout loud enough to have a conversation. They had to speak on the intercom. And most of the time they’d wear earplugs. The B-52 had originally been designed to attack the Soviet Union at an altitude of about 50,000 feet. But Soviet air defenses now forced the bomber to approach at a low altitude — very low. For three to four hours during a training flight, Zink’s plane would fly 150 to 350 feet off the ground. At that altitude, especially in the summer months, the air turbulence was terrible. The hot sun would send thermals of air swirling upward from the ground. Sitting in his little windowless compartment, getting bounced so hard that things would slide off the desk, Zink often felt airsick. But he also felt too busy to get sick. “I’ll throw up later,” he’d tell himself. “I have too much to do right now.”

The navigator would be in constant communication with the pilot, warning of the terrain that was approaching. The B-52’s navigational tools were rudimentary. Its avionics still relied on vacuum tubes, instead of integrated circuits, and data was entered into the bombing computer with IBM punch cards. At low altitudes, the B-52 was an extraordinary sight, a huge plane with a wingspan about sixty yards wide, hugging the terrain, casting a long shadow, traveling seven or eight miles a minute. Zink’s crew often flew through the Rocky Mountains, and the one time that Zink sat in the cockpit, it was fun to watch the pilot bank around hills and drop into alpine valleys. But sitting down below, without any frame of reference other than his radar screen, the experience could be terrifying. On more than one nighttime flight, Zink thought, “we’re going to die,” as the pilot ignored his warning that a mountain was dead ahead and waited an extra moment to climb.

During low-altitude practice runs, Zink’s crew would radar bomb targets throughout the American West, hitting SAC radar huts in places like Sheridan, Wyoming; Bismarck, North Dakota; and La Junta, Colorado. And before a training mission ended, the pilot would spend an hour or two doing “pattern work,” landing the plane, rolling down the runway, and then taking off again. Zink found these touch-and-go landings even harder to endure than heavy turbulence. At the end of every training flight, he felt like someone had just pummeled him for hours.

The Klaxons sounded about once a week during ground alerts. The drills were supposed to be “no-notice” and come as a total surprise. But by the late 1970s, SAC was taking some precautions. Whenever Zink and his buddies saw three fire trucks and the wing commander’s car park on the alert pad, they’d know a drill was about to begin. They’d stand in the tunnel, waiting, making bets on how many seconds would pass before the Klaxons went off. And then they’d run to their planes. As navigator, Zink would decode the message from SAC headquarters. It usually called for an engine start or a “mover,” an exercise that involved taxiing the bomber to the end of the runway, turning around, and returning to the alert pad. Once the drills were completed, the crew would spend about three hours reconfiguring the plane for the next alert.

A few months earlier, during the first week of June, Zink had been fast asleep at about twelve thirty in the morning when the Klaxons sounded. He jumped out of bed, looked out the window — and didn’t see any fire trucks or the wing commander’s car. He and the bombardier thought, “Oh my God, it’s the real thing.” Drills were never held late at night. Hearts pounding, they ran to the plane. Zink decoded the message and felt profoundly relieved that it didn’t contain an emergency war order. The whole episode felt strange, and it wasn’t until weeks later that they learned NORAD had experienced a false alarm. The gunner on Zink’s crew, a young staff sergeant, was so shaken by the experience that he quit the Air Force. All of a sudden, the meaning of their wartime mission had become clear, and he realized, “I can’t do this.” Zink believed strongly in the value of nuclear deterrence and tried not to dwell on what would happen if deterrence failed. He knew that any attack on the Soviet Union by his crew would be not only murderous but suicidal. And yet he never thought about those things while crawling around the Mark 28s and Short-Range Attack Missiles in the bomb bay, checking their serial numbers before an alert.

Zink and his crew were expecting the drill on September 15, 1980. It was about eight thirty in the evening, and out the window you could see the fire trucks and the wing commander’s car. The Klaxons sounded. They ran to the plane. Zink put on his headphones and turned the crew volume low, so he could hear the code from SAC headquarters over the radio.

“Alpha, Charlie, Delta…” he heard, copying each letter down. And then his pilot’s voice was shouting over the intercom.

“Terminate, terminate, terminate.”

For some reason, the pilot was ending the drill. Zink felt scared for a moment, wondering why the pilot was yelling. He and the bombardier looked at each other. They couldn’t see outside, had no idea what was happening — and then heard a loud bang. Something big had struck the right side of the plane. The lights went out, the cabin became pitch black, and Zink knew it was time to evacuate. The navigator was supposed to open the hatch for the rest of the crew and leave the plane first. But the gunner, who sat upstairs, had already jumped down, landed on the floor, and opened the hatch. And without a word, the gunner leaped through the hatch to the tarmac below. Zink’s seat was closest to the hatch, yet four of the five other crew members managed to get out of the plane before him, like rats from a sinking ship. Through the open hatch, Zink could see a bright orange glow — not a good sign.

Zink didn’t bother with the ladder. He jumped the five feet to the runway, landed in a crouch, saw that the right wing of the bomber was on fire, and ran as fast as he could. Now he understood why the crew was in such a hurry. A B-52 had caught fire on the runway a few weeks earlier, at Warner Robins Air Force Base, near Macon, Georgia. Within minutes the plane had exploded, and it literally melted into the ground. But that B-52 hadn’t been carrying nuclear weapons. This one was loaded with eight SRAMs and four Mark 28 bombs.

Zink ran for about three hundred yards, expecting to get knocked down at any second by an explosion. The wing commander’s car pulled up beside him. A window rolled down, and the wing commander said, “Get in.” Zink was glad to obey that order. He turned around and saw that the plane’s number five engine was shooting flames like a blowtorch. It was the engine on the right wing closest to the fuselage, and the fire was cascading down the length of the aircraft. The wing commander was calling firemen on the radio, trying to solve the problem, well aware that not only the plane, but his career at SAC, might be going up in flames.

The nose of the B-52 was pointing toward the southeast, and a wind with gusts of up to thirty-five miles per hour was blowing in that direction. The wind swept from the tail straight down the fuselage, keeping the fire away from the fuel tanks in the wings and away from the bomb bay. Although the power had been shut off on the plane, gravity continued to feed jet fuel into the number five engine. It had become a gigantic flamethrower. Fire trucks sprayed foam on the engine, and yet the steady supply of fuel kept the fire burning. For the moment, the strong wind was pushing the flames away from the B-52. But the wind could change direction, the plane was getting hotter, and its tanks still held another few hundred thousand pounds of fuel.

TIM GRIFFIS WAS AT HOME with his family in Alvarado, Minnesota, a rural town with a population of about four hundred, when the phone rang. Griffis was a civilian fire inspector at Grand Forks Air Force Base, about forty-five miles to the south. His job mainly involved teaching the public about fire hazards and looking at blueprints to make sure that new buildings complied with the fire code. His wife was a schoolteacher at the base. They had a six-year-old son and an eleven-year-old daughter. The kids had gone to bed.

George VanKirk, the fire chief at Grand Forks, was on the phone. The two men were good friends, and they both lived in Alvarado. A B-52 caught fire near the runway about forty minutes ago, VanKirk said. Did Griffis want to come along and help out? Griffis said yes. The two sped to the base as fast as they could in VanKirk’s Ford Fiesta.

By the time Griffis and VanKirk arrived, the fire had been burning for about an hour and a half. The strong wind was still blowing the flames away from the bomber. But the fire trucks couldn’t put out the fire. Some of the hoses were now being used to cool the wings and the fuselage. The copilot had admitted that he might have made a mistake before leaving the plane. Two of the steps in the emergency checklist may have been performed in the wrong order. The checklist said to pull the fire suppression handle for the number five engine, shutting off the fuel — and then turn the emergency battery switch off, cutting the power. The copilot may have turned off the battery first. Without any power, the fire suppression system wouldn’t work, and fuel would continue to flow. Firefighters climbed into the plane twice, entering the cockpit and attempting to perform the steps in the correct order. But nothing happened.

SAC headquarters was on the radio, along with representatives from Boeing, trying to figure out what to do. By quarter to midnight, the fire had been burning for almost three hours. The right wing and the doors of the bomb bay were starting to blister. The fuel tank inside the wing would soon get hot enough to ignite. Boeing’s recommendation was simple: pull the firefighters from the area, abandon the plane, and let it burn. The safety mechanisms on the nuclear weapons would prevent them from detonating, and nobody would get hurt. For some reason, SAC headquarters didn’t seem to like that idea.

VanKirk looked at Griffis and said, “What do you think?”

Griffis knew what the question really meant: somebody should make one last attempt to shut off the fuel.

“Yeah, let me try it,” he replied.

Although Griffis’s current job was fairly sedate, he’d worked for years as a firefighter at Castle Air Force Base in California, where many B-52 pilots were trained. He’d served as the crew chief of a rescue squad, a post that required him to lead men into burning planes as everyone else was leaving them. The interior layout of a B-52 had become awfully familiar, and Griffis thought he could find his way through one blindfolded. But just in case, he wanted Gene Rausch, one of his fire inspectors, to climb into the plane with him — and bring a flashlight.

Their conversation was brief.

“Gene, you want to go with me?”

“Yeah.”

Griffis conferred with the wing commander, going over diagrams of the console and the position of switches in the cockpit. Griffis and Rausch borrowed “silvers,” hooded firefighting suits, from one of the trucks. The boots were two sizes too big for Griffis, and he had to grip the insoles with his toes to walk in them. He stuffed a handheld radio in his hood to communicate with VanKirk, and their conversation was recorded.

“Chief, that engine is getting pretty hot,” Griffis said, five minutes before midnight, “it’s starting to pop, if we’re going to go in, we’ve got to do it now.”

“Yeah, go.”

Griffis and Rausch ran to the plane, entered through the bottom hatch, and climbed into the cockpit. Griffis realized he didn’t need Rausch with him after all. The cockpit was so bright from the flames right outside the window that a flashlight was completely unnecessary. Rausch could have stayed outside in the truck. Griffis had been in burning planes before, but never in one where the fire was cascading with such force. He had no idea if the fuel could be shut off. But he’d give it a try — and if it didn’t work, they’d get their asses out of there. He saw that the fire suppression handle had already been pulled. All he had to do was plug it in. He switched on the emergency battery, and the fire went out, like the burner of a gas cooktop that had just been turned off. And then Griffis and Rausch heard everyone cheering outside.

As Griffis walked from the plane, VanKirk handed him a radio and said, “Here, somebody wants to talk to you.”

It was General Richard Ellis, the commander in chief of the Strategic Air Command.

“Mr. Griffis, I want to thank you,” Ellis said.

Griffis was impressed that the head of SAC knew his name. He subsequently received a Civilian Medal of Valor. But he didn’t consider himself much of a hero. Climbing into a B-52 that was on fire, without power, in the middle of the night, loaded with nuclear weapons, was no big deal. If you’re an Air Force firefighter, he thought, that’s what you do.

During a closed Senate hearing, Dr. Roger Batzel, the director of the Lawrence Livermore National Laboratory, subsequently testified that if the B-52 had caught on fire, the nuclear weapons inside it could have scattered plutonium over sixty square miles of North Dakota and Minnesota. The city of Grand Forks, with a population of about sixty thousand, would have been directly in the path of the radioactive plume. Batzel failed to mention that one of the Mark 28 bombs could have detonated. It would have destroyed Grand Forks and deposited lethal fallout on Duluth, Minnesota, or Minneapolis — Saint Paul, depending on the high-altitude winds. An Air Force investigation discovered the cause of the fire in engine number five: someone had forgotten to screw a nut onto the fuel strainer. The missing nut was smaller than a penny.

Jeffrey Zink and his crew were taken to the hospital, given drug tests, and kept there until three in the morning. They later resented the obsession, among local newspapers, with the question of whether nuclear weapons had been on the plane. The Air Force would neither confirm nor deny it. The crew focused on a more immediate issue: how easily they could have lost their lives. Some of the bombers on alert that night were parked facing west. Had the nose of their B-52 faced west, the fire would have entered the plane the moment the hatch was opened. They would have been incinerated, and the flames would’ve quickly reached the SRAMs and the Mark 28 bombs. The difference between life and death was their parking space.

Not long after the accident, Zink and his wife were having a romantic, candlelit dinner. They were newlyweds. When his napkin brushed the candle and caught on fire, Zink came unglued. All the feelings that had been suppressed hit him at once. He lost it, he felt like a complete basket case. He didn’t have post-traumatic stress disorder or anything really debilitating, just a sudden realization that was hard to express, without sounding trite. Zink was twenty-five years old, and something abstract had become real. These planes are dangerous, he thought. People die in them.

THE DAY AFTER THE B-52 fire at Grand Forks, Senator David Pryor once again introduced an amendment to a Senate bill, calling for the installation of warning sirens at every Titan II launch complex. The commander of the 308th Strategic Missile Wing, Colonel Moser, had informed Pryor that at least nine accidents or propellant leaks had occurred at Titan II missile sites in Arkansas during the previous year. At a launch complex near Heber Springs, a steel rod had fallen onto a circuit breaker, starting a fire and endangering the missile. More than one third of the entire Titan II force had been patched for leaks. Pryor’s amendment was cosponsored by Senator Bob Dole, among others, but it was still opposed by the Air Force. “We have a responsibility to protect the civilians living in the communities and on the farms surrounding these missile sites,” Pryor said during the Senate debate. “Accidents have occurred in the past, and we must take steps to reduce their recurrence and provide for the best course of action in case an accident should occur.”

The Air Force had recently submitted a lengthy report to the House and Senate armed services committees, addressing their concerns about the safety of the Titan II. The report acknowledged that the RFHCO suits and the silo’s communications system could be improved. It also noted that the portable vapor detectors did a poor job of detecting fuel vapor and should be replaced. But the Air Force contended that the accident rate at Titan II sites was lower than the rate at most American workplaces, that current maintenance procedures “provide a high level of safety,” and that the physical condition of the missile was “considered by many to be better now than when it was new.” The safety record of the W-53 warhead was “commendable,” the report said — without mentioning that even the Pentagon thought it needed a retrofit to be safe in abnormal environments. The Air Force argued that the risk of a major propellant leak was low, because the Titan II’s fuel tanks and oxidizer tanks were so well maintained. “Airframe rupture,” the report concluded, “therefore does not constitute a viable concern.”

The Air Force report was useful not only to the Strategic Air Command, which hoped to keep the Titan II on alert, but also to the defense contractors responsible for the missile, like Martin Marietta. They were being sued by Airman Carl Malinger and other victims of the oxidizer leak at Rock, Kansas. But the report didn’t help the Air Force in the Senate. Pryor’s amendment was approved on September 16, 1980, almost a year after it had first been introduced.

Skip Rutherford and his wife were at home, having dinner with an old friend, a couple of days later, when the phone rang. Rutherford got up, took the call, and returned to the table looking white as a ghost.

His wife asked what was wrong.

Somebody dropped a socket in a Titan II silo near Damascus, Rutherford said. The skin of the missile has been pierced, and fuel’s leaking out. The guy who just called says the missile’s going to explode.

Rutherford phoned Senator Pryor, who was in Hot Springs, Arkansas, for the state Democratic convention, along with Governor Bill Clinton and Vice President Walter Mondale.

“This is serious,” Rutherford told the senator.

“Well, how serious?”

“They tell me it’s going to explode.”

“You’re kidding me.”

Outside Rutherford’s house, cars were driving past, kids were playing in yards — and none of them seemed to know that a nuclear disaster might be unfolding, just fifty miles away. Rutherford thought the whole thing was surreal. If the missile did explode, would the warhead detonate? Was the state of Arkansas really about to be wiped off the map? After the conversation with Senator Pryor, the phone at the house kept ringing. The calls were from other staff members, journalists, and the airmen who’d secretly been warning him about the Titan II for months. They said the missile was going to explode, and they hadn’t been wrong yet.

The television was on in the living room, and Rutherford noticed that a good friend of his, Frank Thomas, a twenty-seven-year-old correspondent for Channel 7, was standing across the road from the Titan II site in Damascus. He was repeating the Air Force claim that everything was under control. Rutherford picked up the phone and called Bob Steele, the news director at Channel 7.

“Bob, listen to me,” Rutherford said. “This is totally off the record, but you tell Frank to get the hell out of there.”

“What?”

“Tell Frank to get the hell out of there. He is a friend of mine, and that missile, Bob, is going to explode.”

“How do you know?”

“You have your sources, I have mine,” Rutherford said, beginning to feel a little frantic. “And I’m just telling you, I’m sitting over here watching my friend Frank Thomas on your station standing right before a death trap. And I don’t have a way to get to him, but you all do, and you’ve got to get him out of there.”

Steele got the message. But Thomas was about to leave Damascus, anyway.

Like Hell

The outer door was a real bitch.

The entrance to Launch Complex 374-7 wasn’t protected by high-tech security devices, invented at a top secret weapons lab — just by a heavy steel door, with an electromagnetic lock. And it was hard to open with a crowbar. Greg Devlin and Rex Hukle took turns, one holding the flashlight, the other trying to pry the door open. Nobody had told them how to do it. There wasn’t a checklist for breaking into a Titan II complex, and so the two airmen improvised. They used brute force. Devlin was in pretty good shape from boxing, but the air pack and the RFHCO suit made the work more difficult.

Hukle felt uneasy. They’d walked beneath a thick cloud of fuel vapor to reach the access portal. Now the outer door wouldn’t open. And once they got past this door, they’d have to go downstairs, break open the door in the entrapment area, and open three blast doors with a hand pump to reach the control center. All of that would have to be accomplished within half an hour; their air packs were considered too unreliable after that point. It was about five after two in the morning. They were the only people on the complex. Hukle figured anything could happen and prepared himself for the worst.

Devlin wasn’t having dark thoughts. He just wanted to open the damn door. He felt focused and alert, ready for whatever may come. Devlin’s attitude was: somebody’s got to do this, so it might as well be me.

After fifteen minutes of pulling and prying, the steel door swung open. Devlin and Hukle broke through the entrapment door in about thirty seconds. They left crowbars in both doorjambs to prevent the doors from closing, went down the stairs, and got to work on the first blast door, attaching hoses to its hydraulic valves. Neither of the men had ever used the emergency hand pump before — and the blast door wouldn’t open, no matter how hard they pumped. The fine threads on the hoses were tricky to connect in the dark while wearing rubber gloves. And the pump was an elaborate contraption that didn’t seem to do anything, no matter what they tried. Another fifteen minutes passed, and the blast door was still shut. Their time limit was up. Over the radio, Sergeant Michael Hanson ordered them to quit. Feeling frustrated and defeated, they left the pump beside the door, climbed the stairs, and walked back to the hole in the fence.

Sergeant Hanson, the chief of PTS Team B, led the effort to reenter the control center. He told Devlin and Hukle to read the instructions for the hand pump, grab fresh air packs, go back down there, and try the door again.

Jeff Kennedy thought the whole plan was idiotic. They should be going through the escape hatch, not the access portal. They should have done it at ten o’clock in the evening, not at two in the morning. Almost eight hours had passed since the skin of the missile was pierced. Entering the complex was much more dangerous now — and if something went wrong underground, Devlin and Hukle would be close to the missile, surrounded by fuel vapors, vulnerable to all kinds of bad things.

Let me do it, Kennedy said. I know how to work the pump.

Hanson had tried to send Kennedy back to Little Rock a few hours earlier. He hadn’t asked Kennedy to put on a RFHCO suit; and he hadn’t invited Kennedy to join them at the complex gate. The two men didn’t get along. But Kennedy sure knew a lot about the missile, and he was volunteering.

I’ll go with him, David Livingston said.

Hanson told them to get ready.

While Livingston and Kennedy checked their radios and air packs, Major Wayne Wallace, Sergeant Archie James, and Sergeant Silas Spann left to set up a decontamination area in front of the water treatment building, at the northeast corner of the complex, just outside the fence. When they got to the building, the door was locked, and the combination they’d been given didn’t work. Wallace had to break into the place. Inside, they found a short rubber garden hose. It wasn’t ideal — Livingston and Kennedy would have to walk about a hundred yards to get rinsed off. But it was better than nothing. Spann and James drove over a light-all unit and began to set it up so that the men wouldn’t have to be decontaminated in the dark.

Sergeant Ronald W. Christal showed Livingston and Kennedy the tech order for the emergency hand pump. Christal was a missile pneudraulics technician. He often worked on the blast doors and knew a few tricks to open them that weren’t in the book.

Livingston and Kennedy planned to communicate with each other using hand signals instead of the radios in their RFHCO suits. Only one person at a time could speak on the launch complex radio system — and they wanted to keep the line open as much as possible. One of them would speak to Hanson on the launch complex radio; Hanson would relay the information to Colonel Morris, who’d be right next to him at the pickup truck near the gate. Using the radio in the truck, Morris would speak to Colonel Moser, who was at the command post in Little Rock; Moser would talk to SAC headquarters in Omaha. And, hopefully, as the words passed from one person to another, nothing would be garbled or misunderstood.

At about ten minutes before three, Kennedy and Livingston reached the first blast door. Christal read the instructions for the hand pump to Hanson, who conveyed them over the radio.

The blast door opened.

Livingston took an air sample with a portable vapor detector. They’d been told to check the fuel vapor level every step of the way. If the level exceeded 250 parts per million, they were supposed to leave the complex. The vapor level was 65 ppm in front of the first blast door. As they walked through the door and entered the large blast lock, the level rose to 181 ppm.

At the command post in Little Rock, Sergeant Jimmy D. Wiley heard the vapor level and thought that Kennedy and Livingston should get out of there immediately. Wiley was part of the K crew, the backup team assembled to advise Colonel Moser. Another member of the K crew, Lieutenant David Rathgeber, agreed — if the vapor level was that high after the first blast door, it was bound to be even higher after the second one, as the men got closer to the missile. Wiley and Rathgeber told Colonel Moser that the reentry should be terminated, that the men should be withdrawn from the complex.

The issue was discussed with SAC headquarters. Livingston and Kennedy were ordered to proceed. If they could reach the next blast lock — the small area between the door to the control center and the long cableway to the silo — they could check a panel on the wall that displayed readings from the Mine Safety Appliance. The panel showed the vapor levels in the silo. Kennedy removed the breathing nuts from the second blast door and inserted the probe of the vapor detector through a small hole in the door. Sticking the probe through the door would give a preview of what awaited them on the other side.

The fuel vapor level was about 190 ppm. SAC headquarters told Livingston and Kennedy to open the door, enter the next blast lock, and check the readings on the panel. They opened the door. The room was so full of fuel vapor that they could barely see inside. It looked like a steam room. The portable vapor detector pegged out — the vapor level was far beyond 250 ppm.

Kennedy walked over to the panel. For the first time, he was scared. The blast lock had eight emergency lights, some of them bright red, and he could barely see them. The cloud of fuel vapor floating around them was highly flammable. The slightest spark could ignite it. The RFHCO suits and tools abandoned by PTA Team A were lying on the floor. This is the kind of place you don’t want to be in, Kennedy thought. He looked at the panel, and the needles on the gauges were pointing all the way to the right. They’d pegged out. The gauges said the fuel vapor level in the silo now exceeded 21,000 ppm — high enough to melt their RFHCO suits.

Back out, Hanson said, back out.

Livingston and Kennedy left the blast lock, hurried through the two blast doors, and went up the stairs.

Hanson had an idea: maybe they should turn on a ventilation fan to clear out some of the fuel vapor. The switch for the fan was on the wall of the access portal, at the bottom of the first flight of stairs.

Livingston and Kennedy were almost out of the complex when they heard Hanson say, turn on the fan. They looked at each other. Livingston patted himself on the chest, signaling that he would go down and do it.

Kennedy reached the top of the stairs and stepped into the night air. It felt good to be out of there. That cloud of fuel vapor was insane, he’d never seen anything like it. Kennedy was tired. He decided to sit for a moment on the concrete curb outside the access portal. It had been a hell of a night.

Livingston switched on the fan and came back up the stairs. He was a foot or two behind Kennedy when the Titan II exploded.

At the command post in Little Rock, the radio went dead. And the open phone line from the control center at 4–7 became silent. The sound of the tipsies — the intruder alarm that had been ringing ever since the missile crew left — was gone. Nobody at the launch site could be reached on the radio. For the next eight minutes, the command post did not hear a word from anyone in Damascus. Colonel Moser thought the warhead had detonated.

SID KING AND HIS SALES REP, Tom Phillips, were sitting on the hood of Sheriff Anglin’s squad car, talking with some of the reporters who’d gathered at the access road to the complex, off Highway 65. Nobody seemed worried about the situation. The Air Force had denied there was a serious problem and said everything was under control. But Van Buren County didn’t get a lot of big news stories, and King was willing to hang around a little longer just to see what happened.

A bright white flash lit the sky, and King felt the air around him being sucked toward the missile site. An instant later, a gust blew it back, and a loud sustained roar came from behind the trees, like the sound of a rocket being launched. A column of fire rose hundreds of feet into the air, tall as a skyscraper and towering overhead. The blast briefly turned night into day, pulled the launch complex apart, and lifted the debris into a mushroom cloud. King saw the flames and felt heat on his face and dove to the ground, terrified, as rocks and pieces of concrete began to rain down.

People were screaming, “Get out of here, get out of here,” and a scene that had been calm and quiet a moment earlier became sheer chaos. King and Phillips hid under the taillights of the squad car, trying to avoid falling rocks — and then the taillights came on. Sheriff Anglin was backing out, and he didn’t know they were behind the car. They leaped out of the way as Anglin floored it and pulled onto the road. State officials, highway patrolmen, Air Force officers, cameramen, and reporters were getting into their cars and speeding south toward Damascus. The scene had a primordial feel: every man for himself.

Just a few minutes earlier, Lou Short, the cameraman for Channel 4, had been showing off his brand-new, state-of-the-art, $30,000 RCA video camera. When the debris started to fall, King saw him toss the camera into the back of a truck like an old piece of wood and drive off. None of the news photographers got a picture of the explosion. Getting away from it seemed a lot more important. Larry Ellis, the cameraman for Channel 11, captured the only images of the blast — ten seconds of blurry footage, shot in 16mm, after the eyepiece of his camera was blown off. And ten seconds was long enough for Ellis, who stopped filming, jumped into the truck driven by his reporter, and joined the panicked exodus to Damascus.

King and Phillips headed the other way in the Live Ear, driving north on Highway 65 toward the radio station in Clinton. King had his foot to the floor, praying that his little Dodge Omni could outrun the radioactive fallout and whatever else had been released into the air. They knew the missile had a nuclear warhead, no matter what the Air Force said. But they had no idea if that warhead had detonated. About half a mile up the road, an Air Force security officer stood in the middle of the highway, wearing a gas mask and holding an M-16.

Sheriff Anglin was ahead of them, driving erratically, going ninety miles an hour, slowing down to fifty, and then speeding up again. The cord of Anglin’s police radio had gotten wrapped around his right leg, and every time he lifted the handset to speak into it, the cord pulled his foot off the gas. The sheriff stopped at the truck stop in Bee Branch, about six miles north of Damascus, and told everyone to get out of there, right away. Nobody argued with him. The place emptied, and big eighteen-wheelers peeled onto the highway.

Driving through Choctaw, King realized that neither he, nor Phillips, had spoken since the explosion.

“We just left a bunch of dead people back there,” King said.

“Yeah, I know.”

SAM HUTTO WAS COMING HOME to milk the cows when the missile blew. He was just north of Damascus, on a stretch of Highway 65 that looked down on the launch complex, about two miles away. The blast rattled his pickup. He saw the bright flash, the flames shooting upward like a Roman candle. And then he saw a little sparkly thing fly out of the fire, soar above it briefly, and fall to the ground. He decided not to milk the cows, turned the pickup around, and drove to his brother’s house. And his father, who was spending the night there, seemed curious about what had just happened

“Hop in here,” Hutto said to his dad, “and let’s go up to the top of the hill so you can see.”

Hutto drove to the top of the hill and saw another incredible sight: the headlights of vehicles speeding toward them, bumper to bumper, filling both lanes of the two-lane highway. It looked like a NASCAR restart. As Hutto and his father sat in the pickup beside the road, a couple of state police cars flew past them, followed by news trucks and all sorts of Air Force vehicles — even an ambulance. Nobody stopped, slowed down, or told them to evacuate. Once the vehicles were gone, the road was empty and still again, like it always was at three in the morning. But a fire was burning brightly in the silo.

BOB PEURIFOY WAS FAST ASLEEP when he got the call. A Titan II just blew up in Arkansas, Stan Spray told him. There was a lot of confusion about the details — and no word on the warhead. Of course, it hadn’t detonated full scale. If it had, much of Arkansas would be gone. An Accident Response Group was being assembled, and they wanted Peurifoy to be part of it. A plane would soon land at Kirtland Air Force Base to take him and the rest of the group to Little Rock. Peurifoy got out of bed, thinking about that warhead.

THE DECISION TO EVACUATE the missile site was made by Colonel William Jones, the commander not only of Little Rock Air Force Base but also of its Disaster Response Force. He didn’t have any authority at Launch Complex 374-7 until a disaster occurred. The explosion qualified as one, and Jones briefly conferred with Richard English, the chief of the Disaster Preparedness Division, about what to do. They both thought that everyone at the launch complex was dead and that the air drifting toward Highway 65 was probably full of toxic fumes. Jones knew very little about Titan II missiles and their propellants. He belonged to the Military Airlift Command, which flew transport planes, not the Strategic Air Command.

“Evacuate, evacuate,” English shouted repeatedly, over the loudspeaker of the mobile command post.

Members of the Disaster Response Force were among the first to leave the scene of the disaster.

MICHAEL MAZZARO, the missile crew commander at 4–7, was resting in the back of the ambulance when the missile blew up. Al Childers, the deputy commander, sat beside Ronald Fuller, the missile facilities technician, in the security police pickup that had carried them away from the complex hours before. The truck was parked at the entry control point, about thirty feet down the access road from Highway 65. They were listening to Livingston and Kennedy on the radio. Rodney Holder, the crew’s missile systems analyst technician, was in a truck a few hundred feet closer to the launch complex. None of them thought that the missile was about to explode. Holder hoped that somebody would arrive the next morning, figure out what to do, and fix the problem. They were just sitting there, waiting for that somebody to arrive.

Childers was surprised by the bright white flash. The sun seemed to have appeared in the sky. He knew the warhead hadn’t detonated — and yet, somehow, felt it had.

Fuller opened the door and dove into a ditch.

Holder saw the flash and ducked, heard things hitting the truck, waited a few seconds, took a deep breath, and found it remarkable that he was still alive. Then he got out of the truck and ran toward the highway.

Childers, Holder, and Fuller bumped into one another at the back of the security police pickup. They’d all had the same thought, at the same time: grab the gas mask that you wore out of the launch complex. But the masks were gone, and security police officers were now wearing them. The missile crew members climbed into the backseat of the truck, as the loudspeaker called for everyone to evacuate. Sergeant Thomas Brocksmith, who’d picked them up after they left the control center, got into the driver’s seat. The sky had turned deep red, and Holder worried that a cloud of oxidizer was about to engulf them.

“I need to get the hell out of here before the oxidizer starts falling,” Holder thought. “Everybody at the complex is dead, and Rodney has no need to be here.”

Vehicles were pulling out haphazardly, people were running around in the dark, the evacuation seemed chaotic, and Childers became worried that someone might get hurt. He got out of the pickup truck to direct traffic. It was a thoughtful, well-intended thing to do. Brocksmith drove off without him. Cars and trucks sped past him. Everyone ignored him, and yet somehow nobody got hurt.

About fifty Air Force officers and airmen were at the Titan II site when the missile exploded. Most of them drove to Damascus at high speed. But none of the PTS crew members left. Jim Sandaker had started the night at the barracks, recruiting volunteers to help save the missile. He expressed the PTS point of view, bluntly, when an officer told him to evacuate.

“Screw you,” Sandaker said. “I’m not leaving until I have my friends or their bodies.”

OUTSIDE THE WATER TREATMENT BUILDING, Major Wallace and Sergeant James hid beneath one of the light-all units after the missile blew. As rocks and concrete and little pieces of molten steel landed all around them, James thought: I just want everything to stop falling. The debris lacerated one of his elbows, burned the other, and tore up his left leg. But James was able to stand, and Wallace helped him put on a gas mask. Wallace hadn’t even been scratched. They both wondered what had happened to Silas Spann, who’d been a few feet away from them, seconds before. Now there wasn’t a trace of him.

THE MOMENT THE MISSILE EXPLODED, Silas Spann began to run. He didn’t need to see what was happening — he knew what was happening and instinctively bolted. Hiding under the light-all unit would have made more sense, but running away from the explosion felt a lot better. Spann ran toward the entry control point as fast as he could.

COLONEL MORRIS WAS REACHING for the radio inside his pickup truck when the explosion blew out the windshield. The truck was parked near the gate to the complex and the hole they’d cut in the fence. Morris felt like laughing, as the truck shook and got pounded with debris. It seemed comical: he was sprawled across the front seat and couldn’t get his legs inside the truck, no matter how hard he tried. The door slammed hard onto his left leg. And then he lay there, waiting for something big to hit the pickup. Morris looked up, saw the immense pillar of fire rising from the silo, and put his head back onto the seat.

Hukle was sitting on the tailgate. He managed to crawl across the bed of the truck and hide behind the cab, keeping his eyes tightly shut, amid the roar. Through his eyelids he saw a brilliant red blur. His hands got burned, and something shattered his right kneecap.

Hanson was standing next to the door of the pickup, right beside Colonel Morris. Hanson saw two explosions. The first one shot flames twenty-five feet high out of the exhaust vents, and the second obliterated the silo. The blast wave inflated Hanson’s uniform like a balloon, lifted him off his feet, tossed him down the road, and sent enormous steel beams flying past him.

Christal was standing next to Hanson. He saw the first explosion, missed seeing the second one, flew into the air, and landed twenty feet away. Christal covered his head as the debris fell, got up, looked around, thanked God for sparing his life, and checked to see if all his hair had been burned off. It hadn’t. But the left side of his face and both of his hands were burned.

Greg Devlin was standing about two feet from the gate, facing the silo. The blast wave knocked the wind out of him, like a punch to the stomach, picked him up, threw him onto his back, and slid him fifty feet down the asphalt road. Devlin felt completely under the control of some powerful, malevolent force, unable to move or resist it, propelled by air that seemed to have become rock solid. As Devlin slid down the road on his back, he saw molten steel and pieces of concrete flowing by him like lava.

“Oh shit, you ain’t gonna live through this,” Devlin thought. “I just hope it’s not painful.”

Seconds after the explosion, Devlin was lying in the road, feeling dazed and bleary, like he’d been coldcocked in a boxing match. And then he heard a loud voice in his ear yelling, “Run, run!” The voice scared the shit out of him. He didn’t see anybody, anywhere nearby. Devlin got up, ran for about five steps — and got knocked down again by steel rebar that had just fallen from the sky. It struck his right ankle, tearing the Achilles tendon. The rebar hung from a block of concrete about fifteen feet high and thirteen feet wide. The concrete was part of the silo door abutment. It had landed in the middle of the road, and if he hadn’t gotten up and started to run, it would have landed on him. When Devlin opened his eyes, he saw the shadow of this huge block of concrete, thought the Titan II had landed right next to him, and said to himself, “Oh, my God.”

COLONEL MORRIS LOOKED UP AGAIN, when debris stopped falling onto the truck, and saw that flames were still rising from the silo. He figured it was time to leave. He got out of the truck, and the silhouette of something enormous in the middle of the road made him feel disoriented. Morris heard someone call for help. It was Hukle, sitting in the bed of the pickup, with his RFHCO suit pulled down to his knees. One of the knees was torn open, and Hukle said that he couldn’t walk. Morris pulled the RFHCO suit off him, picked him up, put him over a shoulder, and carried him around the piece of concrete, big as a mobile home, that was blocking the access road.

Devlin saw Colonel Morris and yelled, “Please help, I can’t move.”

Morris carried Hukle for about one hundred yards, lay him down in a field, and then ran back for Devlin. He picked up Devlin and put him over a shoulder.

Devlin could not believe the strength of Colonel Morris. The two were about the same size, and yet Morris was running while carrying him. The man was forty-two years old. Devlin couldn’t stop looking at his face. Blood was pouring down it. Morris looked like he’d been shot in the head.

“I have to put you down,” Morris said. “I have to get to the end of the road, or they’ll leave without us.”

Morris lay Devlin in the field beside Hukle and ran off.

AT THE ACCESS CONTROL POINT, two members of the Disaster Response Force, Richard English and David Rossborough, were preparing to leave for Damascus. Mazzaro was on the radio to the Little Rock command post. Childers was standing nearby when he saw Silas Spann on the access road, sprinting toward them. Spann said there were still people alive on the complex; he’d just helped Hanson and Christal find their way onto the road.

A PTS maintenance officer, Captain George Short, drove up in a station wagon with Colonel Morris, whom he’d seen staggering at the top of the hill.

Colonel Morris wanted to go back for Hukle and Devlin. But Morris looked terrible, and the other men didn’t want him to go. Morris and Short started to do a head count, trying to figure out who was left on the complex. Hanson and Christal appeared, suffering from cuts and burns but strong enough to walk.

Childers grabbed Mazzaro’s gas mask. He didn’t want Mazzaro to be heroic and take any foolish risks — Mazzaro’s wife was about to have a baby. Childers climbed into the station wagon with Rossborough and English. Silas Spann got behind the wheel, and they drove back into the thick of it.

English was the only civilian in the Disaster Response Force. He’d spent twenty years in the Air Force, serving as a navigator in SAC bombers and helping to manage early tests of the Titan II. He retired in 1967, sold insurance for a year, hated it, and got a job at Little Rock Air Force Base that involved a lot of action — training people how to handle disasters, responding to disasters, advising the base commander what to do about disasters. He was fifty-seven, an old man by Air Force standards, and yet greatly admired by his men, who always addressed him as “Colonel.” Far from being over the hill, English was athletic and fit and looked a lot like William Holden, a 1950s movie star.

Rossborough was thirty-two, a sergeant from upstate New York. He’d been at a bowling alley when the Disaster Response Force was recalled. And that explained why, at quarter past three in the morning, on a burning missile site, in the middle of a Broken Arrow, Rossborough was wearing a red bowling shirt.

When the station wagon reached the top of the hill overlooking the complex, the road was littered with debris, and Rossborough told Spann to stop.

The reentry vehicle may have blown apart, Rossborough said, and pieces of the warhead could be scattered everywhere. You don’t want to drive over, or step onto, any of it.

Childers could barely recognize the place that he’d left just hours before. It looked like a war zone. The silo was on fire, the grass was on fire, the hills to the west of the complex and the woods to the north were on fire.

They reached the field where Morris had left Devlin and Hukle. Jim Sandaker was already there, with a fellow member of PTS Team B, Buddy Boylan. They were putting two injured men, Wallace and James, into a pickup truck.

Gene Schneider, another member of the PTS team, had run into the field and picked up Devlin. Schneider carried Devlin in his arms, like a child, as Devlin screamed in pain. His RFHCO suit was dragging along the ground, and every time it got caught on a piece of debris, it applied pressure to his wounded ankle. Schneider would stop for a moment, and Devlin would tell him to keep going. And then Schneider couldn’t carry him any farther. Rossborough and Childers ran over, grabbed Devlin, and placed him in the back of a large truck that Captain Short had driven over. Joseph Tallman, another PTS technician, carried Hukle to the station wagon.

Childers thought he saw the reentry vehicle near the road. Spann was standing right beside it.

“Get away from there,” Childers yelled.

Spann obeyed the order, and when all of the injured had been loaded into vehicles, Childers asked him if anybody was unaccounted for.

Only Livingston and Kennedy, Spann said.

“Let’s go, let’s get out of here,” people shouted.

Nobody was wearing a gas mask. Clouds of oxidizer seemed to be floating above the complex. A large object by the side of the road was loudly hissing; if it was the propane tank, it could explode at any moment. The place did not look or feel remotely safe. Everyone piled into the vehicles, drove off, and returned to the entry control point. Livingston and Kennedy had been left for dead.

COLONEL MORRIS TRIED TO CONTACT the ambulance, using a radio in one of the security police trucks. But the radio in the ambulance was part of the hospital net. It operated on a different frequency. A radio on the security police net couldn’t communicate with a radio on the hospital net. And the radio in the ambulance wasn’t working properly. Captain Donald Mueller — the physician assigned to the Disaster Response Force, who was in the ambulance — could speak to the hospital at Little Rock Air Force Base on the radio. But Mueller couldn’t hear anything that the hospital said in response.

MANY OF THE SECURITY POLICE officers and most of the Disaster Response Force were now in the parking lot of the Sharpe-Payne grocery store in Damascus. It seemed like a good place to regroup. Colonel Jones knew that injured airmen had just been found at the launch complex — but he couldn’t contact the ambulance, either. Speaking to Colonel Morris over the radio, Jones suggested that the injured should be brought to the grocery store.

CAPTAIN SHORT WAS FURIOUS THAT everyone had left the PTS crews at the site, that the ambulance and the security police were nowhere to be seen. Devlin was in great pain. He kept yelling for water, saying his skin was on fire. Devlin’s friends cut the RFHCO suit off him and tried to ease the pain. They didn’t have any painkillers or a medical kit. They emptied a cooler and covered Devlin in water and ice.

“Well, at least I’ve still got the hair on my arms,” Sergeant James said to Childers, “but what’s my face look like?”

Childers thought it wasn’t looking too good. It was burned so badly that most of the skin had peeled away.

Fed up with waiting, Major Wallace said the men should be taken to the nearest hospital. Almost half an hour had passed since the explosion. The injured were placed into a station wagon, a pickup, and a large ton-and-a-half PTS truck. They headed for Damascus.

As the trucks sped south on Highway 65, they passed the ambulance, which was heading north. The PTS truck carrying Devlin and Hukle turned around and drove back to the access road so that a doctor could determine how badly they’d been hurt. Sandaker, driving the pickup, just kept going.

Hukle was put on a stretcher next to the ambulance, and Devlin was examined while lying in the back of the truck. Dr. Mueller thought the injuries didn’t look too serious. But the diagnosis didn’t satisfy Childers or the members of PTS Team B. They took the station wagon and the PTS truck, departed for the hospital in Conway, about twenty-five miles to the south — and, amid the confusion, left Hukle on the stretcher beside the ambulance.

NEAR THE TOWN OF GREENBRIER, about ten miles south of Damascus, Sandaker spotted a couple of security police officers. He stopped the pickup and left two injured men — Hanson and Archie James — with the officers. Then Sandaker did a U-turn and drove north. He wanted to get back to the missile site.

THE HOSPITAL IN CONWAY REFUSED to admit the injured men, claiming that it lacked the authority to treat Air Force personnel. Childers demanded that they be treated and took full responsibility for their care. On the way to the hospital, while sitting in the backseat of the station wagon, Joseph Tallman — the PTS technician who’d carried Hukle from the field — had gone into shock. The refusal to admit these injured young airmen, at four in the morning, about half an hour away from another hospital, seemed in keeping with the spirit of the entire night. The hospital finally agreed to treat them, and Childers called the command post in Little Rock to say where they were.

A FEW HOURS EARLIER, at about one in the morning, after escorting a flatbed truck with light-all units to Launch Complex 374-7, Jimmy Roberts and Don Green had asked if there was anything else they could do to help. They were security police officers with a pickup truck. Devlin and Hukle had not yet broken into the complex with crowbars. Everybody was still waiting for instructions from SAC headquarters.

Sergeant Thomas Brocksmith, the commander of the security police at the site, asked Roberts and Green to drive along the roads surrounding the complex and check on the security officers who were manning the roadblocks. Brocksmith wanted to make sure that all the officers knew how to use their gas masks — in case anything went wrong. Roberts and Green got into their truck and drove along the roads surrounding 4–7. They chatted with security officers at the roadblocks, showing them how to use the masks. Most of the officers didn’t know anything about the Titan II or the danger of its propellants.

At about three o’clock, Roberts and Green were on a road about half a mile southwest of the silo.

The sky lit up.

“Man, ain’t that pretty,” Roberts said, not realizing what had just happened.

A moment later the blast wave shook the pickup so hard it almost went off the road. Roberts and Green quickly put on their gas masks. They had a clear view of the launch complex, and it looked like the fireball extended all the way to Highway 65. They couldn’t reach anybody on the radio and thought that everyone at the complex was dead.

We may be the only two left, Green said.

They decided to evacuate nearby homes — and then heard Sergeant Brocksmith on the radio, calling from the grocery store in Damascus. He told them to evacuate the homes south of the launch complex. They drove east, reached Highway 65, got out of the truck, banged on the doors of small farmhouses and mobile homes, told people to leave at once. Despite the disturbing, early-morning sight of two men in battle fatigues and gas masks standing at the front door, most of the homeowners were grateful for the warning. But one man opened the door, pointed a handgun at them, and said, “I’m not going to leave.” They didn’t argue with him.

Roberts and Green were about a mile north of Damascus when they heard the following exchange over the radio:

“Help! Help me. Help me! Can anybody read me?”

“Yes, we can hear you.”

“Help me!”

“Where are you?”

“This is Sergeant Kennedy.”

“Where are you, Jeff?”

“Colonel Morris, I’m down here by your truck, please help me… my leg’s broke and I’m bleeding.”

“Where are you?”

“I’m down here in your truck!”

Roberts and Green had assumed that they were the only people anywhere near the launch complex. Neither of them had ever met Jeff Kennedy, and they didn’t even know who he was. But they weren’t going to leave him out there. Green turned the pickup truck around and floored it, driving all out, pedal to the metal.

About a minute later, the pickup died right in the middle of Highway 65. It had run out of gas. They got out and pushed it to the side of the road. A passing Air Force truck refused to stop for them, even after they chased it, yelling and waving their arms. The driver of a civilian vehicle swore at them and kept going, when they tried to flag it down. Roberts spotted a Cadillac parked in the driveway of a nearby home, ran over to it, broke one of the windows with a rock, and started to hot-wire the car.

Green was impressed, but not surprised, that Roberts knew how to do that.

A pickup truck approached at high speed from Damascus. Roberts and Green left the Cadillac and stood in the highway, blocking both lanes. They figured: if the truck runs us over, to hell with it.

The truck stopped, and they commandeered it. The driver, Jim Sandaker, insisted on coming with them to the launch complex.

They said, Fine, but get in the backseat.

Green floored it, and the three set out to find Jeff Kennedy.

ONE MOMENT KENNEDY HAD BEEN looking at the ground in front of the access portal, getting ready to sit on the curb. And the next moment he was soaring through the air, spinning head over heels, like an acrobat from a trapeze. And then he blacked out.

When Kennedy opened his eyes, he was lying on his back, and his legs were pointing toward the sky, propped against a chain-link fence. Fires burned all around him. He screamed and yelled for help. But nobody answered.

After lying in that position for a few minutes, wedged against the fence, something inside Kennedy clicked. The choice became clear: he could get up and go — or stay there and die.

Kennedy pulled his legs off the fence, stood up, and immediately fell down. He saw that his right leg was broken, and the rest of him felt bruised and cut up. His helmet was gone. His face was bleeding. After falling down, Kennedy said to himself, “I am not going to die on this complex.”

Using the fence for support, Kennedy pulled himself up and tried to get his bearings. The launch complex was nothing but rubble and flames. It took a little while, but he figured out where he was. The blast had hurled Kennedy about 150 feet through the air. He’d landed upside down against the fence in the southwest corner of the complex. He decided to follow the fence east, toward Highway 65, and then north, hoping to find the hole they’d cut in it. The fence gave Kennedy some physical support and a sense of direction, but it also imprisoned him inside the complex. He couldn’t climb over it, with a broken leg. Until he could find a way out, he was trapped there amid the fires and debris and toxic smoke.

Every few steps, Kennedy fell down. The RFHCO suit was heavy and cumbersome, and without the helmet, it no longer served a useful purpose. It was slowing him down. Kennedy sat on the ground, took off the air pack, and got his arms out of the RFHCO. But he couldn’t pull the suit off his broken leg. He searched the ground, found a jagged piece of metal, and cut the RFHCO suit off above his boots.

Kennedy walked and fell, walked and fell, tripping over debris, looking for the hole in the fence. From somewhere in the darkness, he heard Livingston’s voice, crying out.

“Oh, my God, help me. Please, somebody help me. Please, God, help me.”

“Livy, I’m going for help,” Kennedy shouted.

Livingston didn’t seem to hear him.

“Oh, my God, help me,” Livingston repeated. “Please, somebody help me.”

Kennedy had no idea where Livingston was. The only sign of him was his voice, calling out.

“Please, somebody help me.”

Kennedy kept walking, falling, and getting back up, aware that both of their lives were now at stake. The pain in his leg became excruciating, and he didn’t think he could walk any farther. He started to panic. He thought about his children, his wife. He didn’t want to die on this launch complex. He shouted for help, but nobody answered. And then he told himself to shut up and walk.

In the distance, Kennedy spotted the flashing hazard lights of the pickup truck that Colonel Morris had parked near the gate. The truck was about a hundred yards away — on the other side of the fence. But it gave Kennedy a target, a goal, a destination to reach. Walking and falling, walking and falling, he got close enough to hear chatter on the truck’s radio.

The explosion had knocked over a section of the fence. Kennedy lay on top of it and rolled over it to the other side. He got to the truck and picked up the radio.

COLONEL MORRIS AND CAPTAIN SHORT were sitting in the mobile command post, parked at the end of the access road, talking to Little Rock on the radio. The mobile command post was a pickup truck with two rows of seats in the cab and a camper shell over the back. They both heard a voice on the radio say, “Help,” and then realized it was Kennedy’s.

English and Rossborough jumped into the backseat of the truck, and it took off. Short was driving, Morris giving the directions. He knew exactly where Kennedy was.

The four men in the mobile command post were the last ones at the site who could retrieve Kennedy — and Colonel Morris looked like hell. Dr. Mueller and a medic, Reginald Gray, were in the ambulance on Highway 65, taking care of Hukle. Everyone else was apparently at the grocery store in Damascus, manning roadblocks or en route to the hospital in Conway. English was eager to go back and find this young airman. Rossborough seemed fearless, but this was only his second visit to a Titan II launch complex. His first, about fifteen minutes earlier, had been to rescue Hukle and Devlin.

Short navigated around a deep crater in the road and then stopped the truck. The road was blocked by the slab of concrete that had almost crushed Devlin. They found Kennedy in the battered pickup near the fence and carried him out. He told them that Livingston was still alive, somewhere on the complex, and then asked Short to do him a favor.

“Captain,” Kennedy said, “you have to call my wife.”

Short promised that he would.

Kennedy looked pale. His face was covered with blood. He was having trouble breathing. None of the men were wearing gas masks, and they could smell oxidizer in the air. They had to get Kennedy out of there before searching for Livingston. They lifted Kennedy into the back of the pickup and drove back toward the highway.

A security police truck came toward them on the access road. Short slowed down but didn’t stop. He stuck his head out the window and yelled, we’ve got Kennedy, Livingston is still on the complex, go down there and try to find him.

Roberts and Green had no idea who was in the truck, yelling at them. They didn’t know what Livingston looked like or where he might be. But they were willing to look for him. Green thought about his six-year-old boy, fast asleep at home, completely unaware of what his father was doing right now.

As they neared the complex, a large cylindrical object appeared in the road.

Well, damn, there’s the warhead, Green thought. He carefully drove around it.

Green stopped the truck, and they walked to the northeast section of the complex, looking for a way to get through the fence. They didn’t have a flashlight. Green climbed onto a light-all unit and tried to point it toward the fence, hoping to find a hole. It wouldn’t budge.

The light-all unit was attached to a Dodge Power Wagon, and Green had an idea: I’ll drive this big pickup right through that fence.

Green climbed into the driver’s seat. Someone had left the motor running. He put the engine into first gear and floored it. The truck smashed into the fence, but the fence held. He backed up and tried again — still, no luck. The fence was too strong, and the truck felt kind of sluggish. He got out of the cab and noticed that all four tires had been blown out by the explosion. It was running on rims.

Green thought that Roberts must have returned to their pickup truck. He walked over to it, but nobody was there. He started the truck and followed the southern section of the fence, looking for a hole big enough to drive through. But he couldn’t find one, and the pickup got stuck on some large pieces of cement. After ditching the truck, Green found a small hole in the fence, entered the complex on foot, and started calling for Livingston and Roberts. Nobody replied. It was hard to see anything, with all the smoke and dust. The lenses of his gas mask fogged up. He kept tripping over debris and falling down. He worried that something terrible had happened, that Roberts had fallen into a hole and gotten badly hurt. Green shouted for Livingston and Roberts and realized that he was lost.

JIM SANDAKER HAD BEEN DROPPED OFF at the access control point by the two security officers, and he didn’t plan to remain there for long. The men who’d just returned with Kennedy said that Livingston was still alive but the fumes were pretty strong at the complex. Sandaker looked around for a RFHCO suit, found one, and started to get into it.

Under the Category I rules, you needed at least one other person in RFHCO, as backup, whenever you put on the suit. Colonel Morris objected to Sandaker reentering the complex by himself.

Given the circumstances, Sandaker thought those rules were total bullshit. He was going to look for Livingston.

I’ll go with you, Richard English said, claiming to have been trained to wear the suit.

Sandaker had a feeling that English was lying. He couldn’t believe this old guy was going to put on a RFHCO suit. He worried that English would have a heart attack. Doing anything in a RFHCO was hard work; the whole outfit, with the air pack, weighed almost sixty pounds. The two men had never met, but Sandaker was glad not to be heading into the complex alone.

Colonel Jimmie D. Gray had returned to the site, after looking for water at a nearby farmhouse. Gray had started the night at the Little Rock command post, drove to 4–7 with food and supplies before the explosion, and stuck around after it. He helped Sandaker and English get into the RFHCO suits, and Rossborough drove them to the complex in the mobile command post. This time, he wore a gas mask.

Sandaker and English rode on the back of the truck, dangling their legs over the taillights. Rossborough dropped them off. The communications system on the complex no longer worked, and the two men wouldn’t be able to talk to each other with the headsets inside their helmets. They agreed to signal with their flashlights if one of them got into trouble. They found the hole in the fence and walked through it. From a distance they looked like astronauts exploring a hostile planet.

JIMMY ROBERTS HADN’T SEEN or heard Green slamming the Dodge Power Wagon into the fence. He’d wandered off, searched through Colonel Morris’s battered pickup for a flashlight, failed to find one, and stumbled upon a hole in the fence. Roberts climbed through it and, within minutes, felt completely lost. A couple of thoughts entered his mind: he didn’t want to fall into a hole, and he didn’t want that propane tank, hissing beside the road, to catch on fire and explode. He shouted for Livingston and Green, but got no response. He kept shouting their names — and then he heard someone reply.

“Okay, keep on yelling,” Roberts said, “and I’ll come to your voice.”

About twenty feet from the access portal, Roberts found David Livingston lying on the ground. His face was bloody, and he had a wound in his abdomen. But Livingston was conscious and alert.

Roberts picked him up and started to carry him toward the fence. It wasn’t easy to carry someone while breathing through a gas mask. Roberts started to feel dizzy, and his mask clouded up with sweat.

AT THE ACCESS CONTROL POINT, Don Green suddenly appeared in a pickup truck. Green got out of the truck, looking distraught, and said that Roberts was missing, that he may have fallen into a deep hole. Green needed a new gas mask, he needed to go back to the complex and find Roberts. The others thought Green was delirious, but he felt like they just didn’t understand. His gas mask was clogged, he had to get a new one and find Roberts. Mueller gave Green a shot of Benadryl and persuaded him to sit down for a moment.

WALKING THROUGH THE COMPLEX, Sandaker felt scared. He’d been told to watch out for the warhead and its high explosives. Debris was scattered everywhere, and in the darkness you couldn’t tell what any of it was. The explosion had stripped the concrete off steel rebar, and the rebar had been twisted into all kinds of strange shapes, looming out of the smoke. Sandaker had worked at 4–7 many times, but now nothing seemed familiar. The RFHCO helmet prevented him from calling out for English and Livingston. Within minutes, he was lost.

ROBERTS COULDN’T CARRY LIVINGSTON anymore and put him on the ground.

Livingston pleaded with Roberts not to leave him.

“Look, we’re going to make it out of here,” Roberts said. “I’m going to have to carry you on my back.”

Roberts carried Livingston on his back for a while, but then had to put him down again, unable to carry him another step. Roberts said that he’d go find help and promised to come right back.

“Please don’t leave me,” Livingston said.

Roberts picked him up again and put him on his back.

SANDAKER WANDERED THROUGH THE COMPLEX, looking for the access portal, but couldn’t find it. He felt odd being lost in a place that he knew like the back of his hand. Sandaker spotted English, about thirty feet away. He was turning his flashlight on and off. That meant trouble.

English couldn’t walk any farther in the RFHCO suit. He was exhausted and signaled to Sandaker that he was running out of air.

They turned around and tried to find their way out.

ROBERTS FEARED HE WAS ABOUT to pass out. He put Livingston down near the fence and promised to come back for him. He made his way to the battered pickup near the gate and saw two men in the distance wearing RFHCO suits. He flashed the headlights and honked the horn, but they didn’t see him. And then Roberts saw another truck parked nearby. Someone was sitting in the front seat.

The door of the truck opened, and a man got out, with a flashlight. He was wearing a gas mask and a red bowling shirt.

Roberts thought, “Great.”

Rossborough and Roberts reentered the complex, found Livingston, picked him up, and carried him out. They carried him through bushes and around debris. It felt like running an obstacle course in the dark. They got tired and had to put Livingston down.

As Sandaker and English took off their RFHCOs, they saw Rossborough and Roberts about twenty yards away. They ran over to help, carried Livingston to the truck, and gently lowered him into the back. Sandaker rode with his friend, while the others sat in the cab.

Livingston asked Sandaker not to tell his mother what had happened.

“Please don’t tell my mother,” he said, again and again.

ABOUT AN HOUR after the explosion, Colonel Jones and the rest of the Disaster Response Force returned to the access control point. Jones had been listening to Colonel Morris on the radio and suddenly thought: if he sounds OK back there, what am I doing in Damascus?

Mueller did the best he could to treat Kennedy in the ambulance. Kennedy was pale and thirsty and having difficulty breathing. Mueller started him on an IV and gave him some medicine to prevent pulmonary edema — an excess of fluid in the lungs that could be caused by oxidizer exposure. Kennedy also had a big hole in his right leg. His long johns reeked of rocket fuel, and Mueller cut them off.

Livingston arrived in the back of the pickup, and Mueller examined him there. In some ways, Livingston seemed to be in better shape than Kennedy. His face wasn’t as pale, and he hadn’t passed out. But the wound in his abdomen was deep. Pieces of concrete were lodged in there, and you could see his intestines. Mueller wanted to give him an IV but couldn’t. The ambulance only had one.

Colonel Jones had already requested a helicopter to take Kennedy to the hospital. The command post in Little Rock said that the chopper was on its way. But there was no sign of it.

The helicopter had not yet departed from Little Rock Air Force Base. Its crew had been instructed to bring portable vapor detectors to 4–7. Nobody could find any, and the chopper sat there and waited, for more than half an hour, while people looked for the vapor detectors.

Jones couldn’t understand why the helicopter hadn’t arrived yet. Kennedy and Livingston were in rough shape, and the ambulance wasn’t equipped to deal with their injuries. Livingston needed an IV, right away. Jones told Colonel Morris that he was taking them to the hospital in Conway.

Hukle and Kennedy rode in the ambulance. Livingston remained in the back of the pickup truck with Sandaker, who kept him talking. And Colonel Jones led the way in a station wagon. The convoy had to drive slowly because Livingston was in so much pain.

The helicopter finally took off from Little Rock — without any vapor detectors, because none could be found. The pilot was told to meet the convoy at Launch Complex 374-6, near the town of Republican. But Jones and the others mistakenly drove past it. Instead they met the chopper at Launch Complex 374-5, outside Springhill. The chief of aerospace medicine at the base, another physician, and four medics immediately got to work on the injured men. Kennedy was given a shot of morphine, and Livingston finally got an IV. Sandaker said good-bye to them both, and the helicopter took off for Little Rock. It was five in the morning.

A couple of security police officers picked up Colonel Morris at the access control point and drove him to the hospital.

Colonel Jimmie Gray was the only person left at the site. He waited there, alone, as dawn approached, fires still burned, and the warhead lay somewhere in the dark.

Confirm or Deny

At the Redstone Arsenal in Huntsville, Alabama, Matthew Arnold was taught how to deactivate chemical and biological weapons. “Chlorine is your friend,” the instructor told the class. The principal ingredient in household bleach would render almost every deadly pathogen, nerve agent, and blister agent harmless. That’s good to know, Arnold thought. Although Redstone was an Army facility, he’d been sent there by the Air Force. The three-week course at Redstone was the first step toward becoming an Explosive Ordnance Disposal technician. Students were no longer exposed to nerve gas and then told to inject themselves with atropine — an exercise to build confidence that the antidote would work during a chemical attack. Instead, they were shown footage of a goat being exposed to a nerve agent and given an injection. The goat lived. But the film and the lectures at Redstone suggested how dangerous the work of an EOD technician could be, and a number of people dropped out.

The attrition rate was even higher among those students who, like Arnold, reached the next step — seven months of training, six days a week, at the Naval Explosive Ordnance Disposal School in Indian Head, Maryland. About one third of the students typically flunked out or quit, and only one fifth completed the course on his or her first try. The classes at Indian Head focused mainly on conventional weapons. EOD trainees were required to study every kind of ordnance used by every military in the world. The render safe procedures were similar for most munitions, regardless of their national origin: remove the fuze if it could easily be done, or just attach a small explosive charge to the weapon, retreat a safe distance, and blow it up.

Unlike the bomb squads run by law enforcement agencies, the Air Force EOD teams usually didn’t care about preserving evidence. They were trained to get rid of the hazard, as quickly as possible, and then get out of the way. Arnold learned how to render safe all the conventional warheads, rockets, artillery shells, and bombs in the American arsenal. He also learned how to defuse the sort of handmade, improvised explosive devices used by terrorists groups like the Red Brigades and the Palestine Liberation Front. The handmade stuff could be tricky and unpredictable; the military ordnance, simpler but more powerful. An EOD technician had to approach both kinds with the same mental attitude — disciplined, thoughtful, patient, and calm.

Arnold performed well enough to enter Division Six, the program at Indian Head that taught students how to dismantle a nuclear weapon. The course began with a lesson on the dangers of radioactivity. Every class was shown the film of Louis Slotin dying from radiation sickness in 1946, after his criticality accident at Los Alamos. It was hard to watch. Slotin had been fully conscious and in enormous pain, as his skin swelled, changed color, blistered, and peeled away.

After learning how to use radiation detectors and calculate safe exposure times, the trainees became familiar with various nuclear weapon designs. At the time, the United States had about twenty-five different types — missiles, rockets, warheads, and bombs; artillery shells, depth charges, torpedoes, and mines; large weapons and small ones, atomic and thermonuclear. The most powerful were the Mark 53 bomb, delivered by aircraft, and the W-53 warhead carried by the Titan II. The least powerful was the Mark 54 Special Atomic Demolition Munition (SADM), with a yield of less than 1 kiloton. The SADM weighed only sixty pounds. It was known as a “suitcase bomb” or a “backpack bomb” because of the preferred methods of delivery. One person would carry the SADM and place it in the right spot. Another would set the timer, and then they’d both leave in a hurry.

The instructors at Division Six offered some basic tips on how to deal with a nuclear weapon that’s been in an accident. The first thing you want to do, they said, is find out whether the case of the weapon has been compromised and whether any components have shifted inside it. If your gamma ray detector is showing high levels of radiation, you’ve got a serious problem. Gamma rays will pass right through your protective gear. If you can detect gamma rays from a distance, back away immediately. The weapon may have partially detonated — or it may be about to detonate. But if lives are at stake, calculate how long you can work at the accident site without getting too much gamma radiation.

Always wear a bunny suit, they said, when you walk up to the weapon for the first time. It’s the yellow jumpsuit with the hood. And keep an eye on your alpha and beta meters. If they detect anything, that probably means the weapon’s case has been compromised. The alphas are emitted by the nuclear core, the betas by the tritium gas used to boost it. Your bunny suit will block them, and the respirator will prevent you from inhaling them. And remember: never take off your mask, even if there’s no sign of radiation, until you’re sure that the “skull” of the weapon is intact. The skull is the beryllium reflector around the core. Inhaling beryllium dust can be worse than inhaling plutonium. Both of them can be lethal.

In addition to an alpha meter, a beta meter, a gamma meter, and a tritium meter, an EOD team relied on more prosaic tools to handle a nuclear weapon accident — screwdrivers, ratchets, wrenches, and pliers. The tools were made with metal alloys unlikely to create a spark. If the weapon looked capable of detonating, an EOD technician would open its case with a screwdriver. The most important goal, by far, was to isolate the power sources and ensure that electricity could not reach the detonators. The best way to do that was simply to disconnect the batteries and yank them out. Capacitors that had already charged could be short-circuited with the touch of a screwdriver. But if the X-unit had already charged, an EOD technician had to be very careful. A wrong move could trigger it and detonate the weapon.

Arnold practiced the render safe procedures on dummy weapons that were identical to the real thing — except for the high explosives and fissile material, which were fake. The job was a meticulous process of disassembly. You took the weapon apart, wrapped the parts in plastic, boxed them up, and got them ready for return shipment to the manufacturer. After months of training, Arnold passed all the tests at Indian Head and joined an EOD unit at Barksdale Air Force Base, outside Shreveport, Louisiana. He had learned how to defuse car bombs and biological weapons, to handle Broken Arrows and dismantle nuclear warheads. He was twenty years old.

WHEN SID KING GOT TO CLINTON, he hurried into the radio station and turned on the transmitter. KGFL was licensed to broadcast only during the daylight hours, but the Federal Communications Commission allowed a sunset station to go on the air during an emergency. King thought the explosion of an intercontinental ballistic missile qualified as one. Moments later, his wife arrived at the station, happy to see that he hadn’t been killed. King described the blast to his listeners, and callers to the station shared what they’d seen. Soon the little studio at KGFL was crammed with people, as friends and neighbors gathered there, eager to find out what was going on.

The Air Force was refusing to disclose any information about the explosion. It would not explain what had just happened. It would not discuss the potential danger from toxic fumes. It would neither confirm nor deny that the Titan II was carrying a nuclear warhead. Journalists who called Little Rock Air Force Base were told to phone the headquarters of the Strategic Air Command in Omaha — and nobody at SAC headquarters would answer their questions. SAC headquarters wouldn’t even tell Frank Wilson, the director of environmental services at the Arkansas Department of Health, if the accident had spread radioactive contamination. SAC wouldn’t tell him anything. And so Wilson called a Department of Energy office in Albuquerque. An official there asked him to describe the explosion. Wilson mentioned the fireball and the sparkly thing that seemed to emerge from it. The DOE representative said that the missile probably did carry a nuclear warhead — and it sounded like the high explosives of the weapon had detonated, spreading fissile material. Unable to get any confirmation from SAC, the state of Arkansas sent employees to Van Buren County with radiation detectors.

The Air Force’s silence helped to sow panic and confusion. More than a thousand people left their homes, got into their cars, and fled the area around Damascus. One caller to KGFL said that he was leaving town to stay with family in Fairfield, Illinois, about four hundred miles away. Other callers told of windows being blown out, doors knocked off hinges, an ominous dark cloud that passed over their homes. The cloud smelled like rotten eggs, burned their eyes, and made them cough. The refusal to acknowledge that the missile carried a nuclear weapon made the Air Force seem foolish. One of KGFL’s listeners phoned the station and said that he’d found the radio frequency that SAC was using at the missile site. The conversations with the Little Rock command post weren’t being scrambled. And the whereabouts of “the warhead” were being discussed.

AFTER TELLING THE TRUCKERS in Bee Branch to hit the road, Sheriff Anglin got back into his squad car and drove south on Highway 65 toward Damascus. He wanted to make sure that everybody within five miles of the silo had been evacuated. He stopped at a roadblock north of Launch Complex 374-7. The security police manning it were wearing gas masks.

“Hey, I need one of them masks,” Anglin said.

“Oh, you don’t need a mask,” one of the officers replied, his voice muffled by the mask.

“Well, give me yours, if you don’t need it.”

Neither of them gave Anglin a gas mask, and he headed toward Damascus without one.

The chaos of the early-morning hours extended to the management of roadblocks. The Air Force had no legal authority to decide who could or couldn’t drive on Arkansas roads. But SAC’s failure to confer with state and local officials left a crucial question unanswered — who was in charge? At a roadblock south of 4–7, Air Force security officers refused to let journalists pass. Correspondents from the major television networks had arrived to cover the story, along with radio and newspaper reporters. Sheriff Anglin overruled the Air Force and allowed the media to park on the shoulder of Highway 65, across from the access road to the missile site. It was public property. Not long afterward a reporter for the Arkansas Democrat was stopped at the same roadblock by Air Force security officers and told that he couldn’t drive any farther. The reporter pointed out that his newspaper’s competitors had just been allowed up the road — and then drove around the roadblock without permission and headed toward 4–7, ignoring the soldiers with M-16s. An Air Force security truck pursued him at high speed but gave up the chase. And the correspondent for the Democrat joined the crowd of journalists near the access road, who were shouting questions at every Air Force vehicle that entered or left the site.

AFTER LOADING THE WOUNDED onto helicopters, Richard English and Colonel William Jones returned to 4–7. A convoy from Little Rock Air Force Base soon met them there. It brought specialized equipment and personnel that the Disaster Response Force lacked: portable vapor detectors, radiation detectors, bunny suits, fire trucks, firefighters, and an EOD unit.

A two-man radiation team traveled by helicopter to Launch Complex 374-6 and got a ride from a security police officer to 4–7, about ten miles away. Wearing protective gear, they walked down the access road in the dark, carrying alpha, beta, and gamma ray detectors. They went as far as the low hill overlooking the complex, found no evidence of radioactivity — a good sign — and walked back to the access control point near Highway 65.

English put on a bunny suit and prepared to search for the warhead. The suit was a lot lighter than the RFHCO he’d worn to find Livingston. English thought that he’d seen the warhead during one of his trips onto the complex. His second in command at Disaster Preparedness, Sergeant Franklin Moses, and the members of the EOD unit suited up, too. The half dozen members of the initial reconnaissance team, led by English, waited for permission from SAC headquarters to look for the weapon. The word came from Omaha: they could enter the complex at first light.

RODNEY HOLDER WAS STILL WEARING the T-shirt and old pants he’d put on to take a nap, just before the Klaxons sounded at 4–7. Almost twelve hours had passed since then, and it felt like a long night. Now Holder and Ron Fuller were sitting on the access road to Launch Complex 4–6, outside the town of Republican. They’d hitched a ride from a security police officer at the grocery store in Damascus, hoping to get back to the base in Little Rock. But the officer had gone to 4–6 to pick up a two-man radiation team. And the helicopter had taken off from 4–6 without waiting for Holder and Fuller. The chopper’s departure left them with a couple of options. They could return to the scene of the accident with the radiation team — or stay on the access road at 4–6. The security officer lent Holder his coat and drove off. It was still dark, and the two men sat in the road, exhausted, waiting for someone to give them a ride.

AT THE BAPTIST MEDICAL CENTER in Little Rock, doctors tried to save the lives of Jeff Kennedy and David Livingston. The two were put into the intensive care unit, placed on ventilators, and given high doses of corticosteroids. Oxidizer released by the blast had induced a dangerous form of respiratory distress. Both of the men were now suffering from pulmonary edema, as fluid filled their lungs. Kennedy’s wife left their children with a friend and rushed to the hospital. A young woman came to see Livingston as well, telling one doctor that she was his wife, another that she was his sister. Colonel Michael J. Robertson — the chief of aerospace medicine at the base who’d treated the injured airmen aboard the helicopter — didn’t care who she was. He was just glad that Livingston had someone there. The worst effects of the oxidizer would usually appear about five hours after exposure. Like the phosgene gas used as a chemical weapon during the First World War, the oxidizer could kill you in an extremely unpleasant way. It was known as “dry land drowning.”

MATTHEW ARNOLD HAD BEEN fast asleep when the phone rang at about half past three in the morning. The caller told him to report to the base: his EOD unit was heading into the field. The call came at a bad time. Arnold and his wife had just moved into a new apartment in Shreveport, and they’d stayed up late moving boxes. He’d gotten only a few hours of sleep. The place was full of boxes that still needed to be unpacked, and he didn’t feel like going to work at three in the morning. When Arnold arrived at Barksdale, his squadron commander said that they were going to Arkansas — and nothing more. As the unit loaded its gear into a couple of pickup trucks and prepared to leave, Arnold felt guilty about leaving his wife to deal with the mess at home. He wouldn’t be able to call her, tell her where he was going, or let her know how long he’d be gone.

The EOD team at Barksdale was part of the Strategic Air Command, and it responded to every accident involving SAC nuclear weapons in the eastern half of the United States. During Arnold’s two and a half years as an EOD technician, the unit had spent most of its time on mundane assignments. When unexploded ordnance was found in the marshes surrounding the air base, his EOD squad would defuse it. Every so often, when a plane crashed, they’d render safe the bombs, starter cartridges, flare packages, rounds of ammunition, and ejection-seat rocket motors found in the wreckage. And when nothing else was happening, they’d practice taking apart and reassembling dummy weapons. But on a few occasions, Arnold responded to accidents that involved the real thing.

Twice at Barksdale, a load cart collapsed while transporting a rotary launcher full of Short-Range Attack Missiles. Each launcher held eight SRAMs, and the load carts had telescoping arms to lift the missiles into the bomb bay of a B-52. During both accidents, the telescoping arms broke, dropping the rotary launcher and the SRAMs about five feet to the ground. At least two warheads and half a dozen missiles were damaged. A manufacturing defect or corrosion seemed the most likely explanation for the collapse of the telescoping arms. But an Air Force investigation later found a different cause: maintenance crews had been goofing around with the load carts, out of sheer boredom, and using them to lift B-52 bombers off the ground.

In April 1979, Arnold’s unit responded to a nuclear weapon accident a few miles north of Fort Worth, Texas. The accident was considered serious enough to require their presence urgently, in the middle of the night, and so they flew there in the only aircraft that was available: the base commander’s KC-135. The big jet was a lot plusher than the planes that usually carried Arnold’s team. At Carswell Air Force Base, someone on a loading crew had ignored a tech order and pulled a handle too hard in the cockpit of a B-52. Instead of opening the bomb bay doors, he’d inadvertently released a B-61 hydrogen bomb. It fell about seven feet and hit the runway. When members of the loading crew approached the weapon, they saw that its parachute pack had broken off — and that a red flag had appeared in a little window on the casing. The bomb was armed. Arnold’s team arrived at the base, removed a small panel from the casing, and rotated a switch with a wrench. A green flag replaced the red one in the little window; the bomb was safed. The whole procedure took about an hour, and Arnold’s unit flew back to Barksdale on their usual means of air transportation, a cargo plane.

The prospect of having to render safe a W-53 warhead didn’t make Arnold nervous. The core of the W-53 contained highly enriched uranium, not plutonium, largely eliminating the inhalation hazard and the risk of radioactive contamination. He’d visited Titan II launch complexes, practiced on dummy versions of the weapon. And at Indian Head, he’d been taught that nuclear weapons were almost impossible to detonate accidentally. The safety mechanisms had always worked, even during plane crashes and fires; the high explosives were said to pose the greatest threat to EOD teams. Arnold’s unit handled nuclear weapons all the time, and they rarely thought about the destructive force that could be unleashed. EOD technicians sat on nuclear weapons, casually leaned against them, used them as tables during lunch breaks. But one of Arnold’s commanders was too cocky and nonchalant. He once removed a dummy weapon from a storage bunker in broad daylight, put it into the back of his pickup truck, covered it with a tarp, drove right past security, and disassembled it in front of his girlfriend. Arnold thought the move was stupid and irresponsible, as well as a major breach of security. Inside the bunker, the dummy weapons were stored beside the real ones.

THE RECONNAISSANCE TEAM LEFT the access control point just after sunrise. The search for the warhead didn’t take long. Richard English led them to a spot, about two hundred yards east of the silo, where he thought he’d seen the weapon’s outline in the dark. And there it was, lying in a shallow ditch, right next to the access road. The pickup trucks driving back and forth after the accident, the men stumbling in the darkness, had passed within a few feet of it.

Alpha radiation was detected directly on top of the weapon, but nowhere else on the complex.

The ensuing conversation between Air Force personnel at the site and the Little Rock command post could be overheard by anyone with a shortwave radio:

“It’s laying in a ditch beside, you know, it’s not even up close. It blew out. It’s laying in a ditch. It’s all exposed, and all we need to do is go in and get it.”

“Okay, I’d recommend that we wait for those people that are going to arrive in about an hour.”

“Fine with me.”

IN A HOT SPRINGS HOTEL ROOM, Senator David Pryor and Vice President Walter Mondale were briefed on the Titan II accident. The Democratic Party’s state convention began later that morning, and the vice president was scheduled to give the opening speech. Reporters would be asking questions about the accident, and Mondale wanted to be prepared. Three SAC officers who’d come from Little Rock described the dropping of the socket, the piercing of the missile’s skin, the long wait through the night, the futile attempt to reenter the control center, the explosion. But they refused to disclose whether the Titan II had been carrying a warhead. SAC headquarters had instructed them neither to confirm nor deny the presence of a nuclear weapon.

Mondale picked up the phone and called the secretary of defense, Harold Brown — who wouldn’t tell him, either.

“Goddamn it, Harold, I’m the vice president of the United States,” Mondale said.

Brown told him the missile had a nuclear warhead.

BEFORE TAKING A HELICOPTER to the launch complex, Bob Peurifoy and the rest of the Accident Response Group met with SAC officials at Little Rock Air Force Base. General James E. Light, Jr., had already seen the warhead. Light was the deputy chief of staff for logistics at SAC headquarters. He’d flown to Arkansas, taken a chopper to the accident site, inspected the damage, and returned to the Little Rock command post for this meeting. Light said that the reentry vehicle was remarkably intact, given the size of the explosion.

Peurifoy didn’t like hearing that bit of information. The W-53’s arming and fuzing system, along with its batteries, were attached to the base of the reentry vehicle. And if they were intact, the weapon still had the potential to detonate.

General Light also said that the EOD unit from Barksdale had arrived at the scene, dug a hole under the warhead, wrapped a chain around it, and planned to yank it out of the ditch. But Light had told them not to do anything until the scientists from Los Alamos and Sandia gave their approval.

Peurifoy and William Chambers, who was representing Los Alamos, knew right away that they’d like General Light. He’d made the correct decision. The Air Force was eager to get the weapon out of that ditch as quickly as possible. A crowd of journalists had assembled near the access road on Highway 65, and a small plane carrying a photographer had already flown over the launch complex. But Peurifoy and Chambers thought there was no need to rush the dismantling of America’s most powerful thermonuclear warhead. Chambers knew a fair amount about the subject. He’d responded to the Broken Arrow at Palomares, advised the recovery effort at Thule, written EOD manuals for nuclear weapons, and helped to create the Nuclear Emergency Search Team (NEST), a secretive group that handled threats of nuclear terrorism within the United States. None of the work at Los Alamos and NEST had made Chambers feel anxious — not the weapon accidents, not the ransom note warning of a 20-kiloton bomb in Manhattan, not the warning of a terrorist attack on America’s bicentennial celebrations in 1976. He’d served with General George S. Patton’s Third Army during the Second World War, and the genuine horrors that he saw on the battlefield tempered his fear of hypothetical ones.

From the air, the launch complex looked like it had been hit by a bomb. The Accident Response Group traveled to the site by helicopter, and Peurifoy wondered about the condition of the warhead’s electrical system, his area of expertise. The moment he saw the weapon, Peurifoy thought: Well, my job here is done. General Light had used the wrong term to describe what was lying in the ditch. The “reentry vehicle” wasn’t intact — it was gone, nowhere to be seen, no doubt blown to pieces by the explosion. The warhead lay there by itself, stripped of the electrical power source necessary for a nuclear detonation. But that W-53 was looking pretty good, considering what it had been through. It was still, essentially, in one piece. The outer cover of the primary had been torn off — you could see the detonator cables, the high explosives, the tubing, wiring, capacitors. And the secondary was loose; it no longer sat directly below the primary, like a metal garbage can under a silver basketball. The damage was impressively slight, however, for an object that had flown through a fireball, climbed more than a thousand feet into the air, and hit the ground without a parachute.

Chambers walked over to a nearby Pettibone crane, a mobile all-terrain vehicle that was ready to lift the weapon from the ditch, and drained hydraulic fluid from it. He poured the oil into the holes and cracks of the warhead, coating the high explosives and making them less likely to be set off by a random spark — the kind of spark that might be generated by a chain wrapped around the metal casing of a nuclear warhead. Chambers wanted to know exactly what had happened inside the warhead before anyone tried to move it or take it apart. An X-ray would reveal the amount of damage, but the Accident Response Group hadn’t brought a portable X-ray unit to Arkansas. Until the damage could be properly assessed, Chambers said that nothing should be done. The Air Force had suffered enough embarrassment already. Another accident with this weapon could kill people and spread radioactive tritium, with a pack of journalists literally down the road.

When Matthew Arnold heard about the decision to do nothing until the weapon was X-rayed, he thought it was ridiculous. It was bullshit. It meant his unit would have to sit around in Arkansas for at least another day or two. These civilians don’t know what they’re talking about, he thought. They’re being overly cautious. The warhead doesn’t look that bad, and the render safe procedures wouldn’t be complicated. They were right out of the book.

We’re ready to rock-and-roll, Arnold told his EOD commander. This is what we train for, day after day. These eggheads should just get out of the way and let us go to work. Let’s get the weapon out of here and go home.

Arnold wasn’t allowed anywhere near the warhead. Instead, he was sent onto the launch complex to look for the remnants of retrorockets and other explosives carried by the Titan II. He’d lost his cool, and he knew it. Someone else should do the render safe, Arnold agreed — his mind was too preoccupied with the recent move, the unpacked boxes, the mess waiting for him at home. He wasn’t on top of his game. And much as Arnold hated to admit it, the guy from Los Alamos was probably right.

IN THE PARKING LOT of the hospital at Little Rock Air Force Base, Al Childers was told to take off his clothes. He was contaminated with radiation, according to the alpha detector being used to screen everyone who’d been at 4–7. One of the detectors at the hospital didn’t work at all, and the other kept finding traces of alpha particles. A line of naked men stood in front of Childers, preparing for a rudimentary form of decontamination. They were sprayed with cold water from a garden hose before being allowed into the emergency room. Childers got angry. He’d just come from the hospital in Conway, after making sure that the injured airmen would receive treatment there. He’d pulled a muscle in his back helping to carry Devlin from the field. He didn’t think these alpha readings were accurate. And he couldn’t believe that the hospital was forcing people to strip in the parking lot, while reporters and photographers stood nearby. Childers told the hospital staff to set up a screen or something for a little privacy. This was a harsh welcome home for men who’d had a rough night.

The water from the garden hose felt incredibly cold.

The doctors gave Childers a muscle relaxant for his back and admitted him to the hospital. He was planning to go home later that day, hug his wife, and get some sleep. Instead he was told to get dressed and return to the launch complex. The emergency war order checklists and other classified material had to be retrieved from the safe in the control center. Childers couldn’t understand why his missile crew commander, Mazzaro, hadn’t been asked to do it. But as deputy commander, Childers was responsible for the material, as well. Feeling a bit groggy and wearing the dirty uniform that had just been deemed radioactive, he was driven back to 4–7.

In the early-morning light Childers saw the scale of the destruction for the first time — and realized his life had been spared by sheer luck. The explosion had blown most of the debris toward the west, some of it landing almost half a mile from 4–7. Enormous pieces of steel and concrete lay in the fields of nearby farms. The silo door had been thrown more than two hundred yards, shearing off the tops of trees before crashing into the woods northwest of the complex. The door weighed about 150,000 pounds. Had the debris been blown to the east, toward Highway 65, it would have killed a lot of people.

Driving down the access road, Childers was amused when he saw where the warhead had landed. The object that he’d thought to be the warhead was actually a hydrogen accumulator — a large steel tank, tossed into the road, that looked like the weapon. While yelling for Silas Spann to get away from the tank, afraid that it was the warhead, Childers had been standing right next to the warhead.

Near the entrance to the complex, the road was blocked by debris. Childers and his escorts entered on foot. Smoke still drifted from the silo. The blast had obliterated its upper levels and widened the hole in the ground. What had once been a deep, concrete cylinder now looked like a huge funnel, with a rough edge of rocks and dirt. Security police officers seemed to be everywhere, guarding the site and searching through the wreckage. Childers entered through the access portal, walking down the stairs as Kennedy and Livingston had done earlier that morning. It was dark, and some of the walls and floors were charred. But Childers was impressed that you could still walk through the blast doors and blast locks, that the place was there at all.

The control center felt eerie, like a dark, abandoned basement. Everything was exactly as they’d left it. The Coke that Childers had been drinking was still in its cup. The tech orders and tech manuals were still in their plastic binders, propped open on the floor — none of them had been knocked over by the explosion. The door of the safe was still slightly open, and the classified documents inside it hadn’t moved so much as an inch. Childers and Holder had been right. They’d been right. They could have stayed in the control center. They could have monitored the tank pressures, remained in touch with the command post, turned equipment on or off. And they would have been just fine.

IN THE ABSENCE OF ANY INFORMATION from the Air Force, officials from the Arkansas Department of Health and the Pollution Control and Ecology Department performed their own tests, looking for signs of radiation and oxidizer. About a dozen people in Guy, Arkansas, claimed to have been sickened by toxic fumes. Guy was about six miles from the missile complex. The small town hadn’t been evacuated, and its mayor, Benny Mercer, was among those feeling ill. Everyone seemed to be angry about the federal government’s response. “The Air Force wouldn’t tell us a damn thing when it happened,” a member of the Office of Emergency Services told the Democrat, “and they still won’t.” Gary Gray, the sheriff of nearby Pulaski County, said that he learned more from the radio than from the Air Force. Sam Tatom, the state’s director of public safety, tried to enter the missile site and speak with the commanding officer there, but security police stopped Tatom on the access road, not far from Highway 65.

Governor Bill Clinton found himself in a difficult spot. He had to pacify his own officials, reassure the public, and limit his criticism of the Carter administration, six weeks before the presidential election. After taking a call from Sheriff Gus Anglin, who let him know how poorly everything had been handled, Clinton urged the Air Force to release more details about the accident — and praised its leadership for doing “the best they could.” Vice President Mondale spoke to journalists at the Democratic convention in Hot Springs, accompanied by Governor Clinton, Senator Pryor, and Congressman Bill Alexander. Mondale would neither confirm nor deny the presence of a nuclear warhead. But Alexander was willing to state the obvious. “I assume they’re armed,” he said about the Titan IIs in Arkansas. “That’s why they’re here.”

AT FOUR IN THE AFTERNOON, the secretary of the Air Force, Hans Mark, held a press conference at the Pentagon. Mark was a physicist, a nuclear engineer, and an expert in aerospace technology who’d previously led a research institute at NASA. Mark was the ideal person to explain the inner workings not only of the Titan II but also of the W-53 warhead. He’d been a rocket scientist and a weapon designer. As secretary of the Air Force, Mark provided the Carter administration’s view of the accident.

“I believe that the Titan missile system is a perfectly safe system to operate, just as I believe that the 747 aircraft is a perfectly safe aircraft to operate,” Mark told the press. “Accidents happen.”

When reporters suggested that the Titan II was dangerous, obsolete, and poorly maintained, Mark said that the problem in Damascus hadn’t been caused by equipment failure or a maintenance lapse — it was just an accident, and human error was solely to blame. He refused to answer any questions about the warhead, not even to correct an erroneous claim that plutonium might have been spread by the blast. The explosion was “pretty much the worst case” of what could happen at a Titan II site, he argued. Nobody was killed, no radioactive contamination had occurred, and the only people who got hurt were members of “the emergency teams whose job it is to take these risks.” Unless a more detailed investigation proved otherwise, Mark thought that “the emergency procedures worked properly.”

A COUPLE OF HOURS LATER, David Livingston died at Baptist Medical Center in Little Rock. He’d celebrated his twenty-second birthday the previous week. He was planning to marry his girlfriend in the spring, perhaps leave the Air Force and move to California. She was at the hospital when he passed away; his parents were on an airplane, en route from Ohio, to see him. The official cause of death was pulmonary edema.

Jeff Kennedy remained in the intensive care unit, fighting for every breath.

THE CROWD OF JOURNALISTS in front of the access road swelled on September 20, a full day after the blast. Sid King was impressed by the large truck that a new television network had driven to Damascus. The Cable News Network (CNN) had gone on the air a few months earlier. It was the first television network to offer the news twenty-four hours a day, and the Titan II accident in Damascus was its first big, breaking story. The CNN truck, boasting a huge satellite dish, dwarfed the little Live Ear. CNN correspondent Jim Miklaszewski provided nonstop coverage from the missile site — and broadcast the only images of what appeared to be the warhead, lying on the ground, beneath a blue tarp. To get the shot, Miklaszewski and his cameraman borrowed a cherry picker from a local crew installing phone lines, and rode the cab fifty feet into the air. The Air Force tried, without success, to block their view.

The Titan II explosion fit perfectly with the media narrative inspired by the nuclear accident at Three Mile Island, the taking of American hostages in Iran, and the Carter administration’s failed attempt to rescue those hostages. The United States seemed to have become weak, timid, incompetent. And the “official” version of events was never to be trusted. Although Pentagon rules allowed the disclosure of information about a nuclear weapon after an accident, “as a means of reducing or preventing widespread public alarm,” the Air Force wouldn’t release any details about the warhead in Damascus. When General Lloyd Leavitt threatened to end a press conference in Little Rock if anyone asked another question about the warhead, whose existence had already been televised on CNN, the whole issue became a joke. A newspaper cartoon depicted three Air Force officers: one covering his eyes, one plugging his ears, and one covering his mouth. “If you’re on the military’s side, you can claim that the system worked because the nuclear warhead didn’t go off,” columnist Art Buchwald wrote. “If you live in the area, you may find it hard to sell your house.”

The Soviet Union claimed that the Titan II explosion could have been mistaken for a surprise attack and precipitated “a nuclear conflict.” Senator Pryor and two Republican senators, Bob Dole and Barry Goldwater, demanded a new investigation of the Titan II missile system. “If it’s not safe and effective, I don’t know why you need it,” Dole said.

THE ACCIDENT RESPONSE GROUP EXAMINED the interior of the warhead with the aid of a “pig” — a highly radioactive block of cobalt-60 in a lead box. A sheet of photographic film was placed on one side of the weapon, the pig was put on the other, and the box was opened briefly with a lanyard. Everyone stayed a respectful distance from the pig until the box was shut. The device offered a simple but effective means of taking an X-ray, and it revealed that the warhead was safe to move. Contrary to protocol, the EOD unit from Little Rock was asked to render safe the weapon. Matthew Arnold’s team from Barksdale had to stand and watch as EOD technicians who didn’t even belong to the Strategic Air Command separated the primary from the secondary at 4–7, hidden from CNN’s cameras by a tent. The two sections of the warhead were loaded into separate jet engine containers filled with sand. The containers were lifted onto a flatbed truck, and the truck left the complex as part of a convoy early in the morning on September 22.

“Hey, Colonel, is that what you won’t confirm or deny?” a reporter shouted at one of the passengers, as the truck turned onto Highway 65.

The officer smiled for the cameras and gave a thumbs-up.

The End

Ronald Reagan didn’t feel despair about the future, suffer from a crisis of confidence, or doubt the greatness of the United States. His optimism had tremendous appeal to a nation that seemed in decline. Reagan soundly defeated Jimmy Carter in the presidential election, winning the popular vote by about 10 percent and receiving almost ten times the number of electoral votes. The Republican Party gained control of the Senate and drove four Democratic governors from office — including Bill Clinton, who lost a close race to his conservative opponent. At the age of thirty-four, Clinton became the youngest ex-governor in the United States. The election of 1980 marked a cultural shift, a rejection of liberalism, big government, and the self-critical, apologetic tone that had dominated American foreign policy since the end of the Vietnam War. The new sense of patriotism and nationalism appeared to have an immediate effect. As President Reagan concluded his inaugural address on January 20, 1981, the fifty-two Americans who’d been held hostage for more than a year were released by the government of Iran.

“Peace through strength” had been one of Reagan’s campaign slogans, and his administration soon began the largest peacetime military buildup in the history of the United States. Over the next five years, America’s defense budget would almost double. And the arms race with the Soviet Union would be deliberately accelerated — out of a belief that the United States could win it. Reagan opposed not only détente, but every arms control agreement that the United States had signed with the Soviet Union. In a 1963 speech, he said that President Kennedy’s foreign policy was “motivated by fear of the bomb” and that “in an all-out race our system is stronger, and eventually the enemy gives up the race as a hopeless cause.” The following year Reagan described the Soviets as “the most evil enemy that has ever faced mankind.” His views on the subject remained largely unchanged for the next two decades. He was the first president since Woodrow Wilson who sincerely believed that American military power could bring an end to communism in the Soviet Union.

Most of Reagan’s foreign policy advisers belonged to the Committee on the Present Danger, and they pushed for bold nuclear policies. The counterforce strategy once proposed by Robert McNamara — long associated with RAND and the youthful self-confidence of the early Kennedy administration — was now embraced by conservative Republicans. But the word “counterforce” had become problematic. It sounded aggressive and implied the willingness to fight a nuclear war. Much the same strategy was now called “damage limitation.” By launching a nuclear attack on Soviet military targets, the United States might “limit the damage” to its own territory and, perhaps, emerge victorious.

The new secretary of defense, Caspar “Cap” Weinberger, was, like McNamara, a businessman who’d served in the Army during the Second World War but knew little about nuclear weapons. As a result, his undersecretary of defense for policy, Fred Iklé, played an important role in the Reagan administration’s strategic decisions. Iklé was still haunted by the possibility that deterrence might fail — through an accident, a miscalculation, the actions of a fanatic in the Kremlin. And if that happened, millions of Americans would die. Iklé considered the all-or-nothing philosophy of “assured destruction” to be profoundly immoral, a misnomer more accurately described as “assured genocide.” Aiming nuclear weapons at civilian populations threatened a “form of warfare universally condemned since the Dark Ages — the mass killing of hostages.” He pushed the Reagan administration to seek a nuclear strategy that would deter the Soviets from attacking or blackmailing the United States, maintain the ability to fight a “protracted nuclear war,” limit American damage if that war occurred, and end the war on terms favorable to the United States. A blind faith in mutual deterrence, Iklé believed, was like a declaration of faith during the Portuguese Inquisition — “an auto-da-fé, an act that ends in a mass burning.”

TWO AIR FORCE REPORTS on the Titan II were released to the public in January 1981. One assessed the overall safety of the missile, and the other provided a lengthy account of the accident at Damascus. According to the Eighth Air Force Missile Accident Investigation Board, Launch Complex 374-7 and its Titan II were destroyed by three separate explosions. The first occurred when fuel vapor ignited somewhere inside the complex. The vapor could have been ignited by a spark from an electric motor, by a leak from the stage 1 oxidizer tank, or by the sudden collapse of the missile. A small explosion was followed by a much larger one, as the stage 1 oxidizer tank ruptured, allowing thousands of gallons of fuel and oxidizer to mix. The blast wave from this explosion tore apart the upper half of the silo, tossed the silo door two hundred yards, and launched the second stage of the Titan II into the air. The door was already gone by the time the missile left the silo. The second stage soared straight upward, carrying the warhead, and then briefly flew parallel to the ground. Its rocket engine had been shoved into its fuel tank by the blast. Fuel and oxidizer leaked, causing the third explosion, producing a massive fireball, and hurling the warhead into the ditch.

The accident investigation board determined the sequence of events by examining the fragmentation patterns of the missile and silo debris. Pieces of the second stage were found almost half a mile from the silo, while most of the first stage was scattered within three hundred feet of it. The narrative offered by the report was factual and thorough. But the Air Force seemed more interested in describing how the accident unfolded than in establishing why it happened. “It may not be important whether the immediate cause that initiated the explosive events is precisely known,” the board argued, “since, over a period of time, there were so many potential ignition sources available….”

The Titan II Weapon System Review Group report was prepared for members of Congress. The report contained a number of criticisms and a long list of recommendations for making the missile safer. It said that the vapor detectors in Titan II silos were broken 40 percent of the time, that the portable vapor detectors rarely worked, that the radio system at launch complexes was unreliable and needed to be replaced, that missile combat crews should be discouraged from evacuating the control center during an emergency, that the shortage of RFHCO suits often forced maintenance teams to be selected on the basis of who’d fit into the available suits instead of who knew how to do a particular job, that the suits and helmets were obsolete, that the air packs were obsolete, that some of the missile’s spare parts were either hard to obtain or no longer manufactured, that security police officers should always be provided with maps, that lightning arrestors and other “modern safing features” should be added to the W-53 warhead so that it would meet “modern nuclear safety criteria for abnormal environments.” The report also said that having a warning siren at every launch complex might be useful. The Titan II missile system was “potentially hazardous,” the Air Force concluded, but “basically safe” and “supportable now and in the foreseeable future.”

Jeff Kennedy was angered by both of the reports. He’d spent weeks in the hospital, battling the damage to his respiratory system, and credited a young pulmonologist, Dr. James S. Anderson — not the Air Force — for saving his life. Anderson had sat at Kennedy’s bedside for almost forty hours straight, forcing him to cough up phlegm and clear his lungs. And Anderson had to improvise the treatment for nitrogen tetroxide exposure, since guidance in the medical literature was scarce and nobody from the Air Force would speak to him, for three days after the accident, about the oxidizer or its harmful effects.

The reports were part of a coverup, Kennedy thought: the Air Force cared more about preserving the image of the Titan II missile than protecting the lives of its own men. The accident investigation board said that Kennedy and Livingston were never ordered to turn on the fan in the launch complex. “Do not operate the switch,” Sergeant Michael Hanson told them over the radio, according to the accident report. “Just go to the switch and stand by.”

Kennedy thought the report was wrong. He and Livingston had both heard the order to turn on the fan. Livingston had signaled that he’d go back down and do it; that was one of Kennedy’s last memories before the explosion. Turning on the fan wasn’t part of their original checklist. It was Hanson’s idea. Hanson had suggested it earlier in the evening, while Kennedy and others were arguing that all the electricity should be shut off. And Kennedy had absolutely no doubt that a spark from the fan had caused the explosion. But now Hanson was saying that an order to turn on the fan had never been given, and Colonel Morris was backing Hanson, making the source of ignition seem like some great big mystery. You didn’t need to be a rocket scientist, Kennedy thought, to figure out why the missile exploded. Livingston obeyed the order, turned on the fan — and seconds later the whole place blew up. And the man who was killed by the error was now being blamed for it.

Livingston’s death deeply affected Kennedy. They were close friends, and his death seemed completely unnecessary. Kennedy thought that his commanders at SAC had made a series of mistakes — the decision to evacuate the control center, the refusal to open the silo door and vent the fuel vapor, the endless wait to reenter the complex, the insistence upon using the access portal instead of the escape hatch, the order to turn on the fan. Worst of all was the feeling that he and Livingston had risked their lives for nothing — and then been abandoned. Livingston had lain on the ground for more than an hour, without his helmet, inhaling oxidizer, before anyone came to help. And the delay in sending a helicopter was incomprehensible.

The morale among the PTS crews at Little Rock Air Force Base was terrible. Airman David Powell, who’d dropped the socket that hit the missile, blamed himself for Livingston’s death. A number of PTS technicians refused to work on Titan II missiles, citing the danger of the job, and their security clearances were revoked. Drug and alcohol use increased. The commander of the 308th Strategic Missile Wing, Colonel John Moser, was abruptly reassigned to a desk job at Fort Ritchie in Maryland, overseeing the monthly replacement of computer tapes for the SIOP — a career-ending move. Moser was well liked, and he hadn’t made the crucial decisions that led to the explosion. Nobody at SAC headquarters was fired. Many of the enlisted men in the 308th thought the Air Force was scapegoating the little guys in order to hide problems with the Titan II and protect the top brass.

A few weeks after the accident investigation board’s report was made public, Jeff Kennedy was served with a formal letter of reprimand by the Air Force. It rebuked him for violating the two-man rule and entering the control center at 4–7 without permission. No mention was made of the valuable information he’d obtained there or the bravery he’d displayed trying to save the missile. Air Force regulations permitted a violation of the two-man rule during an emergency, if lives were at risk. But Kennedy wasn’t granted an exemption from the rule. His punishment sent a clear message: the rowdy, hell-raising culture of the PTS crews would no longer be tolerated. They were held responsible for what had gone wrong, not aging equipment or the decisions made at SAC headquarters. And to enforce strict discipline, an officer now accompanied a PTS crew everywhere, like a babysitter, whenever it visited a missile site.

David Powell was given an Article 15 citation — “dereliction of duty” — for attaching the socket to the wrong tool. Powell thought that if he accepted the charge, he’d be admitting negligence and assuming responsibility for the accident. Powell refused to sign and faced the risk of a court-martial instead, where he could defend himself before a panel of military judges. The Air Force didn’t seek a court-martial and gave him a lesser punishment.

Jeff Kennedy had planned to spend the rest of his career in the Strategic Air Command; now he desperately wanted to leave it. Kennedy applied for a medical discharge, hoping to return home and attend college in Maine. The Air Force balked at the request, despite his injuries. Kennedy was sent to Lackland Air Force Base in San Antonio, Texas, for a medical evaluation. He was placed in the psychiatric ward there — along with Greg Devlin, who was also pursuing a medical disability claim.

Devlin had torn his Achilles tendon, suffered burns on his face, neck, back, and hands. He spent ten days at a Little Rock hospital recovering from the skin grafts. But the Air Force was not pleased with Devlin. He’d spoken to reporters about the accident, without SAC’s permission. And he’d filed a $1.5 million lawsuit against the manufacturer of the Titan II, Martin Marietta; members of the armed forces cannot sue the federal government for damages after an injury. David Livingston’s family and Rex Hukle had also decided to sue Martin Marietta. One of the attorneys suing the defense contractor, Bill Carter, was an Air Force veteran and a former Secret Service agent who hoped to obtain compensation for his clients — and to establish in court that the Titan II missile system was unsafe. Carter owned a farm near Damascus and had represented a neighbor sickened by the oxidizer leak there in 1978. During that case, the surgeon general of the Air Force had denied that inhaling oxidizer was bad for you, claiming it was “a substance no more dangerous than smog.”

Devlin could not believe that he and Kennedy had been confined in a mental ward, after everything they’d been through. The place was full of crazy people, like a scene from One Flew Over the Cuckoo’s Nest. Devlin already felt shunned by the Air Force. After returning to duty, he’d been put to work selling hot dogs at the base — a job usually reserved for airmen caught with illegal drugs or facing a dishonorable discharge. But selling hot dogs was preferable to staying in a loony bin. Kennedy would have none of it. He told the staff to release them immediately and move them to a different wing at the hospital — or he’d contact the press. They were promptly transferred. After being examined by physicians, Kennedy was denied a medical discharge, and Devlin was denied a full medical disability. It would have allowed him to use Air Force hospitals for the rest of his life.

A few months later, at a ceremony in Little Rock, both men were given an Airman’s Medal for Heroism, the highest peacetime honor that the Air Force can bestow. Kennedy didn’t want to accept it. But his local congressman in Maine, David Emery, said that if he took the medal, the Air Force would allow him to leave. Kennedy was given the medal by Verne Orr, the secretary of the Air Force, in a room full of reporters. Airman’s Medals were also given to Rex Hukle, Don Green, Jimmy Roberts, and David Livingston’s father. Intended to boost morale, the award ceremony was dismissed by the PTS crews. They thought it was a public relations stunt — and couldn’t understand why Jim Sandaker, who’d returned to the launch complex twice after the accident, didn’t get the highest honor, too.

Jeff Kennedy was granted a “temporary medical leave by reason of disability” three days after receiving his medal. Although the Air Force could recall him to duty in the future, Kennedy’s military career was essentially over. He moved back to Maine, sued Martin Marietta for $7.5 million, and settled out of court for a much smaller sum.

Greg Devlin also left the Air Force within days of receiving the Airman’s Medal. His term of enlistment was over. And his lawsuit was settled out of court, as well. After attorney fees, court costs, and other charges were deducted, Devlin got a check for $6,400.

THE ACCIDENTS AT GRAND FORKS and Damascus had occurred during the same week, and Bob Peurifoy hoped that they would prompt a serious interest in weapon safety at the Pentagon. He traveled to Washington, D.C., and briefed a group of Air Force officials on the design flaws that could detonate a Mark 28 hydrogen bomb during a fire — and the need to retrofit the bombs with new safety mechanisms. The Air Force inspector general and the head of the Air Force Directorate of Nuclear Safety attended the meeting. But it had little effect. A study commissioned by the Air Force later questioned the possibility of an accidental detonation and argued that the Mark 28 didn’t need to be removed from bombers on alert. The study did, however, urge the Air Force to “expedite the proposed retrofit of the 28 and, in the meantime, take extraordinary steps to prevent and ameliorate fires that might involve the unmodified 28s.” Neither of those recommendations was followed.

The Department of Defense had made its spending priorities clear: safety modifications on older weapons like the Mark 28, while desirable, could wait. But Peurifoy was determined to keep fighting the nuclear bureaucracy — and he was willing to engage in a bit of devious behavior, on behalf of weapon safety. After almost twenty years of fierce resistance, the Strategic Air Command had finally agreed to put locks in its bombs. The installation of permissive action links would require new control boxes in the cockpits of SAC’s bombers. Under a contract with the Department of Energy, those new control boxes would be produced by Sandia. Peurifoy quietly arranged for a unique signal generator to be installed in the boxes, along with the coded switch necessary to unlock the PALs. The officials at the Air Force Logistics Command who handled the contract may or may not have understood the purpose of this special, added feature. It allowed all of SAC’s bombers to carry nuclear weapons employing the latest safety devices. The planes would soon be ready — and now Peurifoy had to find a way to get those devices into the weapons.

THE REAGAN ADMINISTRATION’S military buildup was expected to cost approximately $1.5 trillion during its first five years. About $250 billion would be spent on nuclear weapon systems. By the end of the 1980s, the United States would have about fourteen thousand strategic warheads and bombs, an increase of about 60 percent. The Navy would get new cruise missiles and Trident submarines. The Air Force would get new cruise missiles, two new strategic bombers, and one hundred long-range MX missiles, now renamed “the Peacemaker.” The Carter administration’s plan to hide MX missiles amid thousands of square miles in the American Southwest was soon abandoned. Instead, the missiles would be deployed in existing silos — defeating their original purpose and leaving them vulnerable to attack. The only military use of the Peacemaker would be a first strike on the Soviet Union.

The Army’s Pershing II missiles and land-based cruise missiles were among the most controversial weapons proposed by the Reagan administration. They were to be placed in Western Europe, as a counterbalance to the SS-20 missiles recently deployed by the Soviet Union. The SS-20 was not considered a “strategic” weapon — and therefore not covered by existing arms control agreements — because its range was only three thousand miles. An SS-20 missile couldn’t reach targets in the United States. But its three warheads could destroy NATO bases and European cities. The Army’s cruise and Pershing II missiles were intended as a nuclear tit for tat. And yet the Soviet Union considered their deployment extremely provocative. The Pershing II had a range of about a thousand miles and an accuracy of about two hundred feet. From bases in West Germany the Pershing II could destroy command centers in Moscow within five or six minutes. It would give the United States the capability to launch a “super-sudden first strike.”

The new missiles, bombers, and subs gained the most attention in the press. But the “highest priority element” of Reagan’s strategic modernization program was the need to improve the command-and-control system. “This system must be foolproof in case of any foreign attack,” Reagan said. A handful of limited-war options would finally be included in the SIOP, and the ability to fight a protracted nuclear war depended on the survival of command-and-control facilities for days, weeks, or even months. The Pentagon also sought greater “interoperability” — a system that could quickly transmit messages between civilian and military leaders, between the United States and NATO, even between different branches of the American armed services. General Richard Ellis, the head of SAC, told Congress that, at a bare minimum, the command-and-control system had “to recognize that we are under attack, to characterize that attack, get a decision from the President, and disseminate that decision to the forces prior to the first weapon impacting upon the United States.”

The Reagan administration planned to make an unprecedented investment in command and control, spending about $18 billion on new early-warning radars and communications satellites, better protection against nuclear weapon effects and electromagnetic pulse, the creation of a Global Positioning System (GPS) to improve weapon guidance and navigation, upgrades of the bunkers at SAC headquarters in Omaha and at Site R within Raven Rock Mountain, and an expansion of Project ELF, the extremely low frequency radio system for sending an emergency war order message to submarines. Three new ELF antennae would be built in upper Michigan — one of them twenty-eight miles long, the others about fourteen miles long. Project ELF was a scaled-down version of SANGUINE, a plan that had been strongly backed by the Navy. It would have buried six thousand miles of antenna, four to six feet deep, across an area covering almost one third of the state of Wisconsin.

One of the principal goals of the new command-and-control system was to ensure the “continuity of government.” The vice president would assume a larger role in the planning for nuclear war and would be swiftly taken to an undisclosed location at the first sign of a crisis, ready to serve as commander in chief. New hideouts for the nation’s leadership would be built throughout the country. And mobile command centers, housed in tractor-trailer trucks and transported by special cargo planes, would provide a backup to the National Emergency Airborne Command Post.

During the Kennedy administration, the problems with America’s command-and-control system were deliberately hidden from the public. But as President Reagan prepared to adopt an updated version of “flexible response,” the issue of strategic command was discussed in newspapers, books, magazines, and television news reports. Desmond Ball, an Australian academic, made a strong case that a nuclear war might be impossible to control. John D. Steinbruner — who’d helped to write a top secret history of the nuclear arms race for the Pentagon in the 1970s — reached much the same conclusion, warning that a “nuclear decapitation” of America’s leadership could be achieved with as few as fifty warheads. Steinbruner had read the classified studies on decapitation that so alarmed Robert McNamara, but did not mention them in his work. Bruce G. Blair, a former Minuteman officer, described how the command-and-control systems of the United States and the Soviet Union were now poised on a hair trigger, under tremendous pressure to launch on warning if war seemed likely. Paul Bracken, a management expert at Yale University, wrote about how unmanageable a nuclear exchange would be. And Daniel Ford, a former head of the Union of Concerned Scientists, revealed that, among other things, the destruction of a single, innocuous-looking building in Sunnyvale, California, located “within bazooka range” of Highway 101, could disrupt the operation of Air Force early-warning and communications satellites. Although many aspects of Reagan’s strategic modernization program provoked criticism, liberals and conservatives agreed that a robust command-and-control system was essential — to wage nuclear war or to deter it.

In the fall of 1981, Secretary of Defense Weinberger announced the retirement of the Titan II. The missile was increasingly regarded as a relic of another nuclear era. Testifying about the Titan II before the Senate, Fred Iklé cited “its low accuracy and its accident-proneness.” The enormous yield of a single W-53 warhead had become less important. The one hundred Peacemaker missiles scheduled for deployment would carry one thousand warheads — almost twenty times the number carried by the remaining Titan II missiles. And the secrets of the Titan II had recently been compromised. Christopher M. Cooke, a young deputy commander at a Titan II complex in Kansas, had been arrested after making three unauthorized visits and multiple phone calls to the Soviet embassy in Washington, D.C. Inexplicably, Cooke had been allowed to serve as a Titan II officer on alerts for five months after his first contact with the Soviet embassy was detected. An Air Force memo later said the information that Cooke gave the Soviets — about launch codes, attack options, and the missile’s vulnerabilities — was “a major security breach… the worst perhaps in the history of the Air Force.”

Despite the obsolescence of the Titan II, its decommissioning would proceed slowly. The last missile was scheduled to go off alert in 1987. In order to save money, the Air Force decided to cancel some of the modifications recommended by the Titan II review group after the accident in Damascus. Funding would not be provided for a new vapor detection system in the silo, additional video cameras within the complex, or a retrofit of the W-53 warhead with new safety mechanisms. Upgrading the warhead to meet “modern nuclear safety criteria for abnormal environments” would have cost about $400,000 per missile.

THE SOVIET INVASION OF AFGHANISTAN, the breakdown in détente, the tough rhetoric from the White House, and the impending arrival of cruise missiles and Pershing II missiles created widespread fear of nuclear war in Western Europe. The fear was encouraged by a Soviet propaganda campaign that sought to stop the deployment of America’s new missiles. But the apocalyptic mood in Europe was real, not Communist inspired, and loose talk by members of the Reagan administration helped to strengthen it. Thomas K. Jones, an undersecretary of defense, played down the number of casualties that a nuclear war might cause, arguing that families would survive if they dug a hole, covered it with a couple of doors, and put three feet of dirt on top. “It’s the dirt that does it,” Jones explained. “Everyone’s going to make it if there are enough shovels to go around.”

In Great Britain, membership in the Campaign for Nuclear Disarmament soon increased tenfold. A quarter of a million CND supporters attended a demonstration in London’s Hyde Park during the fall of 1981, and a well-publicized Women’s Peace Camp grew outside the Royal Air Force Base Greenham Common, where American cruise missiles would soon be housed. In Bonn, a demonstration against the Pershing II missile also attracted a quarter of a million people. The sense of powerlessness and dread, the need to take some sort of action and halt the arms race, led to a nuclear version of the Stockholm syndrome. Throughout Western Europe, protesters condemned American missiles that hadn’t yet arrived — not the hundreds of new Soviet missiles already aimed at them.

The New Yorker magazine ran a three-part article in February 1982 that catalyzed the antinuclear movement in the United States. Written by Jonathan Schell and later published as a book, The Fate of the Earth revived the notion that nuclear weapons confronted the world with a stark, existential choice: life or death. Schell tried to pierce the sense of denial that had seemingly gripped the United States since Hiroshima and Nagasaki, the refusal to face the threat of annihilation. “On the one hand, we returned to business as usual, as though everything remained as it always had been,” Schell wrote. “On the other hand, we began to assemble the stockpiles that could blow this supposedly unaltered existence sky-high at any second.” He called for the abolition of nuclear weapons, offered a chilling description of what a single hydrogen bomb would do to New York City, and presented the latest scientific evidence on how nuclear detonations could harm the ozone layer of the earth’s atmosphere. Later that year the astronomer Carl Sagan conjured an even worse environmental disaster: nuclear winter. The vast amount of soot produced by burning cities would circle the earth after a nuclear exchange, block the sun, and precipitate a new ice age. Sagan warned that the effects of nuclear winter would make victory in a nuclear war impossible; a nation that launched a first strike would be committing suicide.

On June 12, 1982, perhaps three quarters of a million people gathered in New York’s Central Park, demanding a different kind of freeze — a worldwide halt to the production of nuclear weapons. The New York Times called it “the largest political demonstration in American history.” The Nuclear Weapons Freeze Campaign gained the support of mainstream groups like the U.S. Conference of Mayors, the National Council of Churches, and the Roman Catholic Church. Unlike the European antinuclear movement, it called upon both the United States and the Soviet Union to disarm. But the campaign threatened the Reagan administration’s strategic modernization plans, and opponents of the freeze claimed that it was being orchestrated by “KGB leaders” and “Marxist leaning 60’s leftovers.” By the end of 1982, about 70 percent of the American people supported a nuclear freeze. And more than half worried that Reagan might involve the United States in a nuclear war.

NINETEEN EIGHTY-THREE PROVED to be one of the most dangerous years of the Cold War. The new leader of the Soviet Union, Yuri Andropov, was old, paranoid, physically ill, and staunchly anti-American. A former head of the KGB, Andropov had for many years played a leading role in the suppression of dissent throughout the Soviet bloc. The election of Ronald Reagan persuaded him that the United States might seek to launch a first strike. The KGB began an intensive, worldwide effort to detect American preparations for a surprise attack, code-named Operation RYAN. Andropov’s concerns were heightened by the Reagan administration’s top secret psychological warfare program, designed to spook and confuse the Kremlin. American naval exercises were staged without warning near important military bases along the Soviet coastline; SAC bombers entered Soviet airspace and then left it, testing the air defenses. The Soviet Union played its own version of the game, keeping half a dozen ballistic-missile submarines off the coast of the United States.

On March 8, 1983, at the annual convention of the National Association of Evangelicals, President Reagan called the Soviet Union “the focus of evil in the modern world… an evil empire.” Two weeks later, Reagan announced his Strategic Defense Initiative, soon known as Star Wars, a long-range plan to defend the United States by shooting down enemy missiles from outer space. The technology necessary for such a system did not yet exist — and Reagan acknowledged that it might not exist for another ten or twenty years. But Star Wars deepened the Kremlin’s fears of a first strike. An American missile defense system was unlikely to be effective against an all-out Soviet attack. It might, however, prove useful in destroying any Soviet missiles that survived an American first strike. Andropov strongly criticized the plan and warned that it would start a new arms race. “Engaging in this is not just irresponsible,” Andropov said. “It is insane.”

The Pershing II missiles were supposed to arrive in West Germany at the end of November, and anxieties about nuclear war increased throughout Europe as the date approached. On the evening of September 1, Soviet fighter planes shot down a civilian airliner, Korean Airlines Flight 007, killing all 269 of its passengers. The Boeing 747 had accidentally strayed into Soviet airspace, not far from a missile test site, and the airliner was mistaken for an American reconnaissance plane. The Kremlin denied that it had anything to do with the tragedy — until the United States released audio recordings of Soviet pilots being ordered to shoot down the plane. President Reagan called the attack “an act of barbarism” and a “crime against humanity [that] must never be forgotten.”

A few weeks later alarms went off in an air defense bunker south of Moscow. A Soviet early-warning satellite had detected five Minuteman missiles approaching from the United States. The commanding officer on duty, Lieutenant Colonel Stanislav Petrov, tried to make sense of the warning. An American first strike would surely involve more than five missiles — but perhaps this was merely the first wave. The Soviet general staff was alerted, and it was Petrov’s job to advise them whether the missile attack was real. Any retaliation would have to be ordered soon. Petrov decided it was a false alarm. An investigation later found that the missile launches spotted by the Soviet satellite were actually rays of sunlight reflected off clouds.

During the third week of October, two million people in Europe joined protests against the introduction of Pershing II missiles — and a team of Army Rangers, Navy Seals, and U.S. Marines led an invasion of Grenada, a small island in the Caribbean. The invasion had ostensibly been launched to protect the lives of American citizens and restore order amid the aftermath of a military coup. It also achieved another goal: the overthrow of a Communist regime backed by the Soviet Union and Cuba. Nineteen American soldiers, twenty-five Cubans, and forty-five Grenadians were killed in the fighting. The Soviet Union condemned Operation Urgent Fury as a violation of international law. But it was enormously popular in the United States, boosting President Reagan’s image as a strong, decisive leader. A long time had passed since Americans had been able to celebrate a military victory.

The invasion of Grenada, however, revealed a number of serious problems with the World Wide Military Command and Control System. The Army’s radio equipment proved to be incompatible with that of the Navy and the Marines. According to a Pentagon report, at one point during the fighting, unable to contact the Navy for fire support, “a frustrated Army officer used his AT&T credit card on an ordinary pay telephone to call Ft. Bragg, NC [the headquarters of the 82nd Airborne Division] to have them relay his request.”

The week after the invasion, NATO staged a command-and-control exercise, Able Archer 83. It included a practice drill for NATO’s defense ministers, simulating the procedures to authorize the use of nuclear weapons. The KGB thought that Able Archer 83 might be a cover for a surprise attack on the Soviet Union. The timing of such an attack — a few weeks before the arrival of the Pershing IIs — seemed illogical. Nevertheless, “the KGB concluded that American forces had been placed on alert,” a Soviet agent later wrote, “and might even have begun the countdown to war.” A number of the Soviet Union’s own war plans called for using military exercises as a cover for a surprise attack on Western Europe. While NATO played its war game, Soviet aircraft in Poland and East Germany prepared to counterattack. Able Archer 83 ended uneventfully on November 11 — and NATO’s defense ministers were totally unaware that their command-and-control drill had been mistaken for the start of a third world war.

On the evening of November 20, American fears of nuclear war reached their peak, as ABC broadcast The Day After, a made-for-television movie. Directed by Nicholas Meyer, starring Jason Robards, and set in Lawrence, Kansas, the film combined melodrama with a calm, almost documentary account of how the world might end in 1983. Some of the most powerful images in The Day After had nothing to do with mushroom clouds, radiation sickness, or the rubble of a major American city. When Minuteman missiles first appear above Kansas, launched from rural silos there and rising in the sky, the film conveyed the mundane terror of nuclear war, the knowledge that annihilation could come at any time, in the midst of an otherwise ordinary day. People look up, see the missiles departing, realize what’s about to happen, and yet are powerless to stop it. About 100 million Americans watched The Day After, roughly half of the adult population of the United States. And unlike most made-for-television movies, it did not have a happy ending.

THE PERSHING II MISSILES ARRIVED in West Germany, and the Soviet Union’s response was purely diplomatic. Its negotiators walked out of arms control talks and didn’t return. The relationship between the two superpowers had reached its lowest point since the dangerous events of 1962. And while billions of dollars were being spent on new strategic weapons in the United States, the safety problems with older ones continued to go unaddressed. Earlier in the year, another B-52 had caught on fire on a runway at Grand Forks Air Force Base. It was undergoing a routine maintenance check, at 9:30 in the morning, when fuel suddenly ignited, created a large fireball, destroyed the plane, and killed five young maintenance workers. No nuclear weapons were involved in the accident. But similar B-52s were being loaded with Mark 28 bombs and Short-Range Attack Missiles every day.

A program to add new safety devices to the Mark 28 — weak links and strong links and a unique signal switch — was begun in 1984. But the retrofits were halted a year later, because the program ran out of money. Thousands of the bombs remained unmodified. And the safety problems with the Short-Range Attack Missile were worse than originally thought. The high explosives used in the primary of the SRAM were found to be vulnerable to fire. As the missiles aged, they also became more hazardous. The propellant used by their rocket motors had to be surrounded at all times by a blanket of nitrogen gas. When the gas leaked, the propellant became a “contact-sensitive explosive” that could easily be set off by flames, static electricity, or physical shock. If the SRAMs were poorly maintained, simply dropping them on the ground from a height of five or six feet could make them explode — or take off. “The worst probable consequence of continuous degradation… is spontaneous ignition of the propellant in a way similar to a normally initiated burn,” an Air Force nuclear safety journal warned. “Naturally, this would be a catastrophe.” The journal advised its readers to “follow procedures and give the weapons a little extra care and respect.”

Bill Stevens retired from Sandia in 1985. His job had been redefined during a management shake-up, and he lacked enthusiasm for bureaucratic infighting. He was disappointed that most of the weapons in the stockpile still didn’t possess the safety devices his team had pioneered. But Stevens felt proud of his recent contribution to the safety of the Pershing II. Hoping to eliminate human error during launch exercises with the missile, the Army had decided to computerize the procedure. At Pershing II bases in West Germany, crews would install the warhead, erect the missile, remove the pin that locked the missile onto its launcher, run the countdown until one second before launch — and then stop the exercise. The countdown would be controlled by a computer. Stevens felt uncomfortable with the idea; in fact, he thought it was crazy. A software glitch could launch a Pershing II missile. And the Army’s software, written in 1980, was unlikely to be bug free.

Stevens refused to sign off on the nuclear weapon system study of the Pershing II missile, citing the risk of a DUL — a deliberate, unauthorized launch. In response to his criticisms, a safety device was added to the first-stage rocket motor. It required a separate code, entered manually, before the missile could take off. The warhead atop the Pershing II contained a permissive action link and wouldn’t have detonated after an accidental launch. But the Soviet Union wouldn’t have known that fact, as the missile on their radar screens headed toward Moscow.

RONALD REAGAN, despite all his tough rhetoric, had long harbored a fear of nuclear war. His first years in the White House increased that fear. During a command-and-control exercise in March 1982, Reagan watched red dots spreading across a map of the United States on the wall of the Situation Room. Each dot represented the impact of a Soviet warhead. Within an hour the map was covered in red. Reagan was shaken by the drill and by how little could be done to protect America. Although some members of the administration viewed the Strategic Defense Initiative as a clever response to the growing antinuclear movement, an attempt to show America’s aims were peaceful and defensive, Reagan’s belief in the plan was sincere. He thought that a missile defense system might work, that it could save lives, promote world peace, render nuclear weapons “impotent and obsolete.” Reagan had a sunny, cheerful disposition, but watching The Day After left even him feeling depressed. With strong encouragement from his wife, Nancy, he publicly called for the abolition of nuclear weapons. Reagan’s criticism of the Soviet Union became less severe, and his speeches soon included this heartfelt sentiment: “A nuclear war cannot be won and must never be fought.”

The deaths of Yuri Andropov and his successor, Konstantin Chernenko, brought Mikhail Gorbachev to power. Gorbachev represented a dramatic break from the past. He was youthful and dynamic, the first Soviet leader since Vladimir Lenin who’d attended a university. Although Gorbachev’s attempts to change the Soviet Union were tentative at first, he was committed to reforming its stagnant economy, allowing freedom of speech and religion, ending the war in Afghanistan, rejecting the use of force against other nations, linking the Soviet bloc more closely to the rest of Europe, and abandoning the pursuit of nuclear superiority. Although many of his views were radical, compared to those of his predecessors, Gorbachev did not seek to betray the tenets of Marxism-Leninism. He hoped to fulfill them.

In age, temperament, background, education, political orientation, Gorbachev and Reagan could hardly have been more different. And yet they were both self-confident, transformational leaders, willing to defy expectations and challenge the status quo. During their first meeting, at a Geneva summit conference in November 1985, the two men established a personal rapport and discussed how to reduce the nuclear arsenals of both nations. Gorbachev left Geneva viewing Reagan not as a right-wing caricature, a puppet of the military-industrial complex, but as a human being who seemed eager to avoid a nuclear war.

A year later, at a summit in Reykjavik, Iceland, the discussion strayed onto a topic that alarmed many of Reagan’s close advisers: huge reductions in the number of nuclear weapons. Secretary of State George P. Shultz was elated by the possibility. The recent accident at the Chernobyl nuclear power plant had deposited radioactive fallout across much of Europe and the Soviet Union, reminding the world of the far greater danger that nuclear weapons posed. Reagan and Gorbachev seemed on the verge of reaching an extraordinary agreement, as a transcript of their meeting shows:

The euphoria that Reagan and Shultz felt didn’t last long. Moments later Gorbachev insisted, as part of the deal, that all Star Wars testing must be confined to the laboratory. Reagan couldn’t comprehend why a missile-defense system intended to spare lives — one that didn’t even exist yet, that might never exist — could stand in the way of eliminating nuclear weapons forever. He refused to place limits on the Strategic Defense Initiative and promised to share its technology. The Soviet Union was conducting exactly the same research, he pointed out, and an antiballistic missile system had already been built to defend Moscow. Neither Gorbachev nor Reagan would budge from his position, and the meeting ended.

Despite the failure to reach an agreement on the abolition of nuclear weapons, the Reykjavik summit marked a turning point in the Cold War, the start of a process that soon led to the removal of all intermediate-range missiles from Europe and large cuts in the number of strategic weapons. The all-out nuclear arms race was over. Gorbachev now felt emboldened to pursue reform in the Soviet Union, confident that the United States did not seek to attack his country. And the hard-liners in the Reagan administration breathed a sigh of relief, amazed that their president had come so close to getting rid of America’s nuclear weapons. Margaret Thatcher, the conservative prime minister of Great Britain, and François Mitterrand, the socialist president of France, were furious that Reagan had questioned the value of nuclear deterrence, a strategy that had kept the peace since the Second World War. Although European protest marches had focused mainly on the United States for the previous six years, it was the leadership of Western Europe who most strongly opposed creating a world without nuclear weapons.

BOB PEURIFOY HAD BECOME a vice president at Sandia, and his new status enabled him to lobby more effectively for nuclear weapon safety. By 1988 almost half of the weapons in the American stockpile were fitted with weak link/strong link devices, and the safety retrofit of Mark 28 bombs had finally resumed. But SAC was still loading about one thousand Short-Range Attack Missiles onto its bombers on alert. Those planes were parked on runways nationwide, ready to take off from bases in California, Kansas, Maine, Michigan, New Hampshire, New York, North Dakota, South Dakota, Texas, and Washington State. As tensions between the United States and the Soviet Union eased, the Air Force’s willingness to risk an accident with a SRAM became harder to justify.

On February 26, 1988, Peurifoy wrote to the assistant secretary for defense programs at the Department of Energy and invited him to Sandia for a briefing on the dangers of the SRAM. The assistant secretary never replied to the letter. The following month, the president of Sandia raised the issue with another official at the DOE, who suggested that the secretary of energy and the secretary of defense should be briefed on the matter. But nothing was done. A few months later an independent panel was commissioned to look at management practices at the Department of Energy, and Peurifoy was asked to serve as a technical adviser. Headed by Gordon Moe, a former member of Henry Kissinger’s national security staff, the panel wound up using the SRAM’s safety problems as a case study in mismanagement. Moe was shocked by the lack of attention to nuclear weapon safety and its implications. Almost fifteen years had passed since concerns about the SRAM were first expressed — and yet no remedial action had been taken. “The potential for a nuclear weapon accident will remain unacceptably high until the issues that have been raised are resolved,” the Moe panel said in a classified report. “It would be hard to overstate the consequences that a serious accident could have for national security.”

John H. Glenn, a former astronaut and a Democratic senator from Ohio, visited Sandia on April 26, 1989. Peurifoy took the opportunity to give Glenn a briefing on nuclear weapon safety — and handed him a copy of the Moe panel’s report. Glenn wanted to know more about the subject and asked whom he should contact at the Department of Energy to discuss it.

Peurifoy suggested that he skip the midlevel bureaucrats and raise the issue with the secretary of energy, James D. Watkins.

Glenn said that he’d be seeing Watkins the following week.

The bureaucratic logjam was broken. A well-respected senator — a national hero — planned to raise the issue of nuclear weapon safety with someone who could actually do something about it.

Secretary Watkins and his staff met with Senator Glenn, read the Moe panel report, got worried about the safety of older weapons in the stockpile, and contacted the secretary of defense, Dick Cheney, about the issue. Instead of taking the weapons off alert, the Pentagon commissioned two more studies of the SRAM. One would be conducted by the Air Force, the other by Gordon Moe — who was rehired by the Department of Energy to repeat his earlier work.

Almost another year passed. The Berlin Wall had fallen. Mikhail Gorbachev had visited the White House; signed major arms agreements; removed hundreds of thousands of Soviet troops from Eastern Europe; allowed Poland, Hungary, Czechoslovakia, East Germany, Romania, Latvia, Estonia, and Lithuania to leave the Soviet bloc. By any rational measure, the Cold War was over. But every day, across the United States, Short-Range Attack Missiles continued to be loaded into B-52s on ground alerts.

During the spring of 1990, R. Jeffrey Smith, a reporter at the Washington Post, learned about the safety problems with some American nuclear weapons. The Post ran a series of his articles, bringing public attention to the SRAM’s flaws and to the W-79 atomic artillery shells’ lack of one-point safety. Smith didn’t divulge any classified information, but he did suggest that bureaucratic rivalries and inertia were creating unnecessary risks. A Pentagon spokesman defended the SRAM, claiming that the “weapon meets all our current safety standards.” Secretary of Defense Cheney met with Air Force officials, Secretary of Energy Watkins, the heads of the three weapons laboratories, and General Colin Powell, the chairman of the Joint Chiefs of Staff, to discuss the SRAM. On June 8, 1990, Cheney said that the SRAMs posed “no safety hazards to the public” — but that they would immediately be removed from bombers on alert, until another safety study was completed.

The House Armed Services Committee had already appointed a panel of three eminent physicists to investigate the safety of America’s nuclear weapons. Charles H. Townes was a Nobel laureate who had advised the Department of Defense for many years. John S. Foster, Jr., was a former director of the Lawrence Livermore Laboratory who’d served in high-level posts at the Pentagon during the Johnson and Nixon administrations — an expert not only on nuclear weapon technology but also on targeting strategies. Sidney Drell, the chairman of the panel, was a theoretical physicist, long associated with the Stanford Linear Accelerator, who for many years had served as a JASON — a civilian granted a high-level security clearance to help with sensitive defense matters. Drell, Foster, and Townes didn’t always agree on nuclear weapon policies. Drell had opposed the MX missile; Foster had supported it. But they shared a mutual respect, and their expertise in the field was unsurpassed. Peurifoy was asked to serve as a technical adviser.

The Drell Panel on Nuclear Weapons Safety submitted its report to the House Armed Services Committee in December 1990. The report confirmed what Bill Stevens and Bob Peurifoy had been saying for almost twenty years: America’s nuclear arsenal was not as safe as it should be. Recent improvements in computing power, the report noted, had led to “a realization that unintended nuclear detonations present a greater risk than previously estimated (and believed) for some of the warheads in the stockpile.” The Drell panel recommended that every nuclear weapon should be equipped with weak link/strong link devices, that every weapon carried by an airplane should contain insensitive high explosives and fire-resistant nuclear cores — and that the Pentagon should “affirm enhanced safety as the top priority of the U.S. nuclear weapons program.”

A separate study on nuclear weapon safety was requested by the House Foreign Affairs Committee. The study was conducted by Ray E. Kidder, a Lawrence Livermore physicist, and released in 1991. It gave a safety “grade” to each nuclear weapon in the American stockpile. The grades were based on their potential risk of accidental detonation or plutonium scattering. Three weapons received an A. Seven received a B. Two received a C plus. Four received a C. Two received a C minus. And twelve received a D, the lowest grade.

ON JANUARY 25, 1991, General George Lee Butler became the head of the Strategic Air Command. During his first week on the job, Butler asked the Joint Strategic Target Planning Staff to give him a copy of the SIOP. General Colin Powell and Secretary of Defense Dick Cheney had made clear that the United States needed to change its targeting policy, now that the Cold War was over. As part of that administrative process, Butler decided to look at every single target in the SIOP, and for weeks he carefully scrutinized the thousands of desired ground zeros. He found bridges and railways and roads in the middle of nowhere targeted with multiple warheads, to assure their destruction. Hundreds of nuclear warheads would hit Moscow — dozens of them aimed at a single radar installation outside the city. During his previous job working for the Joint Chiefs, Butler had dealt with targeting issues and the damage criteria for nuclear weapons. He was hardly naive. But the days and weeks spent going through the SIOP, page by page, deeply affected him.

For more than forty years, efforts to tame the SIOP, to limit it, reduce it, make it appear logical and reasonable, had failed. “With the possible exception of the Soviet nuclear war plan, this was the single most absurd and irresponsible document I had ever reviewed in my life,” General Butler later recalled. “I came to fully appreciate the truth… we escaped the Cold War without a nuclear holocaust by some combination of skill, luck, and divine intervention, and I suspect the latter in greatest proportion.”

Butler eliminated about 75 percent of the targets in the SIOP, introduced a targeting philosophy that was truly flexible, and decided to get rid of the name SIOP. The United States no longer had a single, integrated war plan. Butler preferred a new title for the diverse range of nuclear options: National Strategic Response Plans.

MIKHAIL GORBACHEV WAS ON VACATION in the Crimea on August 18, 1991, when a group calling itself the “State Committee for the State of Emergency” entered his house and insisted that he declare martial law or resign. After refusing to do either, Gorbachev was held hostage, and the communications lines to his dacha were shut down by the KGB. His military aides, carrying the nuclear codes and the Soviet equivalent of a “football,” were staying at a guesthouse nearby. Their equipment stopped functioning — and the civilian leadership of the Soviet Union lost control of its nuclear weapons.

Two other Soviet officials possessed nuclear codes and footballs: the minister of defense and the chief of the general staff. Both of them supported the coup d’état. It has never been conclusively established who controlled the thousands of nuclear weapons in the Soviet arsenal during the next few days. The head of the air force later claimed that he, the head of the navy, and the head of the Strategic Rocket Forces took over the command-and-control system, preventing anyone else from launching missiles at the United States. After the coup failed on August 21, communications were restored to Gorbachev’s dacha, and the football carried by his military aides became operable once again.

Eager to reduce the risk of an accidental war and encourage deeper cuts in the Soviet arsenal, President George H. W. Bush announced a month later that the United States would unilaterally make large reductions in its nuclear deployments. It would remove all of the Army’s tactical weapons from Europe, destroy half of the Navy’s tactical weapons and place the rest in storage, take 450 Minuteman II missiles off alert — and end the Strategic Air Command’s ground alert. For the first time since 1957, SAC’s bombers wouldn’t be parked near runways, loaded with fuel and hydrogen bombs, as their crews waited for the sound of Klaxons.

The Soviet Union ceased to exist on Christmas Day, 1991. The following June, the Strategic Air Command disappeared, as well. General Powell and General Butler thought that SAC had outlived its original purpose. The recent war against Iraq had demonstrated the importance of close collaboration between the armed services — and future wars were likely to be fought with conventional, not nuclear, weapons. The Strategic Air Command and its institutional culture no longer seemed relevant. SAC’s aircraft were divided among various Air Force units. America’s land-based missiles and ballistic-missile submarines were assigned to a single, unified command — to be headed, alternately, by an officer from the Air Force or the Navy. The fierce interservice rivalry to control America’s nuclear weapons largely vanished, as those weapons played an increasingly minor role in the Pentagon’s war plans. But many SAC veterans were outraged that what had once been the most powerful organization in the American military was being disbanded. They thought it was a mistake, regarded General Butler as a turncoat, and felt that the legacy of Curtis LeMay was being dishonored.

President Bush told members of his administration not to brag or gloat about the downfall of the Soviet Union, an event with myriad causes that Mikhail Gorbachev had unintentionally but peacefully overseen. General Colin Powell ignored those instructions at the ceremony in Omaha marking the end of the Strategic Air Command. “The long bitter years of the Cold War are over,” Powell said. “America and her allies have won — totally, decisively, overwhelmingly.”