A few minutes past midnight on 24 January 1961, a Mark 39 hydrogen bomb drifted down through the winter darkness over Wayne County, North Carolina, swinging beneath a 30-metre (100-foot) parachute. It settled nose-first into a field near the hamlet of Faro and came to rest almost upright, its canopy snagged in a tree. The weapon's design yield is commonly cited at 3.8 megatons — roughly 250 times the destructive power of the Hiroshima bomb.

A second, identical bomb was falling at the same time. Its parachute never opened. It slammed into a muddy farm field at a speed reportedly around 1,100 km/h (700 mph) and shattered, burying parts of itself deeper than anyone would ever manage to dig. The aircraft that had carried both — a B-52G on a Cold War airborne-alert mission — was tumbling out of the sky in pieces.

For decades the official line was that there had never been any real danger. Then, in 2013, journalist Eric Schlosser pried a 1969 Sandia National Laboratories memo loose through the Freedom of Information Act while researching his book Command and Control. Its verdict on that night in North Carolina fit into a single sentence — and it did not agree.

Eight Men and a Leaking Wing

The B-52G had launched from Seymour Johnson Air Force Base on an airborne-alert mission, part of the Strategic Air Command routine that kept nuclear-armed bombers aloft around the clock. Around midnight the crew rendezvoused with a tanker for aerial refuelling. The tanker crew noticed something the bomber's instruments had not yet made obvious: fuel was streaming from the B-52's right wing.

The refuelling was aborted. Within minutes the leak went from worrying to catastrophic — the aircraft commander, Major Walter Scott Tulloch, reported that roughly 17,000 kg (37,000 lb) of fuel had escaped in just three minutes. Ground control ordered the bomber back to Seymour Johnson. It never made the runway.

Descending through 10,000 ft (3,000 m), the pilots lost control. Tulloch ordered the crew out at 9,000 ft (2,700 m). Five men ejected or bailed out and landed safely; one ejected but did not survive the landing; two died in the crash. Third pilot Lt. Adam Mattocks scrambled out through the top hatch without an ejection seat — he is the only person known to have escaped a B-52 that way and lived. The empty bomber broke apart in mid-air, scattering wreckage across some five square kilometres (two square miles) of tobacco and cotton farmland.

Two Bombs, Two Very Different Falls

As the airframe disintegrated, both Mark 39s tore free. Weapon no. 1, riding in the aft bomb bay, was flung out at around 9,000 ft. The violent break-up did something the crew had never commanded: it yanked the safing pins from the bomb's arming rods exactly as a deliberate release would have, and the retard parachute deployed as designed.

Weapon no. 2 left the forward bay much lower, somewhere between 5,000 and 2,000 ft (1,500–600 m). Its parachute never opened. The bomb hit a waterlogged field at free-fall speed and broke apart, driving its components deep into the mud.

Recovery teams found weapon no. 1 standing almost upright, parachute draped in a tree, with what the official assessment called negligible damage. The unnerving part was what the weapon had been doing on the way down. The extracted arming rods had fired its pulse generator, which activated the low-voltage battery; the 42-second timer ran to completion; and the barometric pressure switch closed as the bomb descended — step after step of a live drop sequence.

Popular accounts, following the documents Schlosser obtained, usually summarize it this way: three of the bomb's four safeguards failed or were bypassed. Sandia's own analyses framed the details slightly differently over the years, but the core finding is not in dispute — by the time the bomb reached the ground, only one device in the firing chain still said no.

The Switch That Held

That device was the MC-772 arm/safe switch, a solenoid-operated electro-mechanical switch buried in the bomb's electronics. It stayed on SAFE unless a deliberate 28-volt signal from the cockpit moved it to ARM. On weapon no. 1 it remained on SAFE, and the high-voltage firing circuit was never completed.

Later reviews made plain how thin that margin was. Jones concluded in 1969 that the Mk 39 Mod 2, in his words, did not possess adequate safety for the airborne alert role in the B-52 — a stray electrical short during a mid-air break-up, he judged, could credibly have thrown the switch. A 1987 Sandia study raised the same worry. Sandia safety engineer Bill Stevens later summed up the ambiguity: “Some people can say, ‘hey, the bomb worked exactly like designed.’ Others can say, ‘all but one switch operated, and that one switch prevented the nuclear detonation.’”

The Bomb That Would Not Come Out

Weapon no. 2 was a different kind of problem. It had buried itself in a swampy field, and explosive ordnance disposal teams spent days digging through freezing mud and groundwater just to reach its components. It took three days to recover the one part everyone wanted to see: the arm/safe switch.

Lt. Jack ReVelle, the EOD officer who led the recovery, described the moment his team finally pulled that switch out of the crater. It is one of the most quoted exchanges of the entire Cold War, and he repeated it in interviews for the rest of his life.

The finding was less clear-cut than the exchange suggests — the immediate technical analysis recorded the mangled switch as electrically neither in the armed nor the safe position, and this bomb had not charged its firing circuit. But nobody standing at the edge of that crater found the distinction comforting.

The weapon's primary stage came out on 30 January from about 6 m (20 ft) down, its high explosives crumbled but undetonated. The thermonuclear secondary kept sinking out of reach. With the crater flooding faster than the pumps could clear it, the Air Force abandoned the dig at 42 ft (13 m). University of North Carolina researchers later estimated the secondary sits about 55 m (180 ft) below the field, where it remains today — the government purchased an easement over the site that restricts any digging.

From Denial to Declassification

For half a century, official statements emphasized that the safety systems worked. The 1969 Jones memo, published by The Guardian in September 2013 after Schlosser's FOIA request, showed that the people responsible for weapon safety had privately concluded something closer to the opposite. The document became a centrepiece of Command and Control, Schlosser's history of American nuclear accidents, later adapted into a PBS documentary.

Goldsboro pushed weapon designers toward the modern doctrine of nuclear surety — environmental sensing devices, stronger electrical isolation, and safeguards that assume the aircraft itself may come apart. The lesson was reinforced five years later, when another B-52 broke up during refuelling and dropped four hydrogen bombs near Palomares, Spain — two of which scattered plutonium across the countryside.

North Carolina eventually acknowledged its brush with catastrophe in its own understated way. In 2012 the state erected a highway marker in the nearby town of Eureka. It reads, simply, “Nuclear Mishap.”

Had the switch on weapon no. 1 been jolted to ARM, a contact-fuzed thermonuclear detonation in eastern North Carolina was — in Sandia's own later language — credible. One simple, low-voltage switch said no.

Sources: Wikipedia, The Guardian, National Security Archive, Eric Schlosser – Command and Control, ECU News, Sandia National Laboratories declassified documents