A Routine Flight Goes Wrong
At 12:33 a.m. on January 24, 1961, a B-52G Stratofortress — tail number 58-0187, assigned to the 4241st Strategic Wing — departed Seymour Johnson Air Force Base on a routine airborne alert mission designated "Cover All 22." The aircraft was part of Strategic Air Command's Operation Chrome Dome, a continuous nuclear deterrence patrol that kept nuclear-armed bombers aloft 24 hours a day along predetermined routes, ready to strike Soviet targets on presidential order [1]. The crew of eight was commanded by Major Walter Scott Tulloch.
During an aerial refueling rendezvous with a KC-135 tanker at approximately 10,000 feet over eastern North Carolina, the tanker crew noticed fuel streaming from the B-52's right wing. They radioed a warning. Within minutes, the leak had become catastrophic — the bomber had lost over 17,000 kilograms (37,000 pounds) of fuel from a ruptured bladder cell in the right wing [1]. Major Tulloch was instructed to hold in a wide orbit near Seymour Johnson while emergency procedures were coordinated, but the massive structural stress from the asymmetric fuel loss proved too much. At approximately 12:33 a.m., twelve miles north of Goldsboro near the rural community of Faro, the right wing failed catastrophically. The aircraft broke apart in midair at an altitude between 2,000 and 10,000 feet [2].
As the bomber disintegrated, it released its payload: two Mark 39 Mod 2 thermonuclear bombs, each carrying a yield of 3.8 megatons — roughly 250 times more powerful than the weapon dropped on Hiroshima and capable of producing a fireball hot enough to vaporize steel [2]. Five of the eight crew members survived by ejecting or being thrown clear in the breakup. Three died — one man's parachute was caught in the wreckage, another was trapped in the cockpit, and a third landed in a tree but died of exposure before rescuers could reach him in the swampy terrain.
But the real story was on the ground.
Two Bombs, Two Fates
The first bomb descended on its parachute exactly as designed — as if it had been deliberately released over a target during a combat mission. The weapon's retarding system activated in sequence: the drogue chute deployed to stabilize the bomb's fall, followed by the main 100-foot-diameter parachute that slowed it to a survivable descent rate [1]. It drifted to earth in a tobacco field near the community of Faro, landing upright and largely intact.
When the recovery team reached it, they found something terrifying: of the four arming mechanisms in the weapon, three had activated during the fall. The bomb's spin motors had started — the centrifugal arming devices that normally require deliberate pilot action. Its barometric pressure switches had closed at the proper altitude, interpreting the descent as a legitimate weapons delivery. Its trigger mechanism had engaged, and the piezoelectric firing signal was ready [2]. Only a single low-voltage switch — a simple, dynamo-technology safety device that was supposed to receive an electrical signal from the cockpit before enabling detonation — remained in the "safe" position. It was the last barrier between a quiet January night in rural North Carolina and a thermonuclear explosion.
Lieutenant Jack ReVelle, the 25-year-old bomb disposal officer from the Explosive Ordnance Disposal team at Seymour Johnson who led the recovery, arrived at the scene within hours. He later told journalists and documentary filmmakers: "How close was it? My opinion is we came damn close. When you have a weapon that has the potential for a 3.8-megaton yield, and one switch prevented a full-yield thermonuclear explosion — that's too close" [3]. ReVelle described the experience of approaching the intact bomb, hanging from its parachute lines tangled in a tree, as the most terrifying moment of his military career.
The second bomb followed a very different trajectory. Its parachute failed to deploy — likely because the breakup forces severed the deployment lanyard — and it plummeted into a muddy, waterlogged agricultural field at approximately 700 miles per hour [1]. The tremendous impact drove the weapon deep into the soft, swampy Carolina soil. The force of the collision shattered the bomb's casing and scattered its components — including radioactive uranium and tritium — through the surrounding earth to depths exceeding 180 feet in some projections.
The Air Force dispatched heavy excavation equipment and spent weeks attempting recovery. Workers dug a crater roughly 30 feet deep and pumped thousands of gallons of groundwater before recovering most of the weapon's conventional explosive components, the tail assembly, and portions of the casing [1]. But the nose cone and the uranium "pit" — the weapon's fissile core, containing both enriched uranium and plutonium — had sunk beyond reach into the saturated soil. The water table made deeper excavation impossible without risking collapse and contamination of the local aquifer.
The Cover-Up
The Air Force acknowledged the crash publicly — it was impossible to hide a B-52 disintegrating over a populated area, raining debris across several square miles of farmland — but systematically downplayed the danger. Official statements issued through the Pentagon's public affairs office said the bombs had "multiple safety devices" that prevented any possibility of detonation, and that at no time was there any risk to the public [4]. The term "Broken Arrow" (the military code for a nuclear weapons accident involving potential or actual detonation, nuclear contamination, or loss of a weapon) was used internally but not shared with the public or local officials.
Local newspapers reported the crash as an aviation accident. The Goldsboro News-Argus covered it as a B-52 that experienced "structural failure," noting the deaths and the recovery of "components." No mention was made of how close the weapon came to functioning as designed [3]. Residents of Faro and surrounding communities were not evacuated, not warned of radiological contamination from the shattered second bomb, and not informed that a thermonuclear weapon had nearly detonated above their homes.
For over fifty years, the Air Force maintained that there had never been any realistic danger of detonation. This narrative held through every congressional inquiry and journalist's question until 2013, when investigative reporter Eric Schlosser obtained declassified documents through the Freedom of Information Act while researching his book *Command and Control: Nuclear Weapons, the Damascus Accident, and the Illusion of Safety* [3]. Among those documents was a formerly secret report by Sandia National Laboratories engineer Parker F. Jones, written just eight days after the accident. Jones had been tasked with conducting an internal safety assessment. His conclusion was unambiguous: "One simple, dynamo-technology, low-voltage switch stood between the United States and a major catastrophe" [2].
Jones further noted that the switch in question was known to be vulnerable to short-circuiting under certain conditions, and that the same type of switch had been found in the "armed" position in other accidents. His memo recommended urgent redesign — a recommendation that was classified and kept from the public for five decades [2].
What a Detonation Would Have Meant
Had the single remaining switch failed — through short-circuiting, impact damage, or the electrical transients common during aircraft breakup — the 3.8-megaton weapon would have detonated at or near ground level. To understand the scale: 3.8 megatons is equivalent to 3,800,000 tons of TNT. The atomic bomb that destroyed Hiroshima was 15 kilotons. The Goldsboro weapon was over 250 times more powerful [2].
The resulting fireball would have been over a mile in diameter, with core temperatures exceeding 100 million degrees Fahrenheit — hotter than the interior of the sun. Everything within the fireball's radius would have been vaporized instantly: soil, structures, vegetation, and any living thing. The blast wave — a wall of compressed air traveling at supersonic speeds — would have leveled every structure within a 5-mile radius, including much of Goldsboro, a city of roughly 30,000 people in 1961 [2].
The thermal radiation pulse would have caused third-degree burns — the full-thickness destruction of skin — at distances up to 12 miles from ground zero. Fires would have ignited spontaneously across an area of nearly 500 square miles. The detonation crater would have been approximately 400 feet deep and a quarter-mile across [3].
But the most catastrophic long-term effect would have been fallout. A ground-level detonation of a 3.8-megaton weapon would have scooped millions of tons of irradiated soil into its mushroom cloud, which would have risen to altitudes above 50,000 feet. Depending on prevailing wind patterns — which in January typically blow from southwest to northeast along the Eastern Seaboard — lethal fallout could have drifted over Raleigh, Richmond, Washington D.C., Baltimore, Philadelphia, and New York City within 24 hours [3]. Conservative estimates from nuclear weapons effects calculators suggest potential casualties in the hundreds of thousands from blast and thermal effects alone, with fallout-related deaths and cancers potentially numbering in the millions over subsequent years.
The geopolitical consequences are almost impossible to calculate. In January 1961 — with the Cold War at its peak, the Berlin Crisis building, and John F. Kennedy just four days from inauguration — a nuclear detonation on American soil could have been misinterpreted by Soviet early-warning systems as a first strike or could have triggered panic-driven escalation [4].
The Buried Bomb
The second bomb remains in the ground to this day. After weeks of excavation in early 1961, the Air Force recovered most of the weapon's components from depths of up to 20 feet — including the thermonuclear "secondary" stage, the conventional explosive lenses, and portions of the arming system. But the uranium and plutonium core was never found [1]. The nuclear material — estimated at several kilograms of enriched uranium-235 and plutonium-239, with half-lives of 700 million years and 24,000 years respectively — remains embedded somewhere in the waterlogged earth of Wayne County.
The military purchased a permanent easement on the affected land, fencing off a small circular area roughly 200 feet in diameter. The Air Force Corps of Engineers installed monitoring wells around the perimeter and conducts periodic radiological sampling of the groundwater [1]. Results have consistently shown no significant migration of radioactive material into the water table, though critics note that the monitoring is infrequent and the contamination plume may take decades to reach detectable levels in sampling wells.
Local residents were told almost nothing. The farmer whose land contained the buried nuclear material — Adam Mattocks's family property — learned of the full nature of what lay beneath his fields only years after the fact [3]. A small historical marker was eventually placed near the site, but for decades there was no public acknowledgment of what remained underground.
Systemic Failures
The Goldsboro incident was not an isolated event. Between 1950 and 1968, the United States experienced at least 32 officially acknowledged Broken Arrow incidents — accidents involving nuclear weapons that resulted in launching, firing, detonation, theft or loss of the weapon, or radioactive contamination [4]. These included bombs accidentally dropped from aircraft over New Mexico, South Carolina, and Spain; weapons involved in fires on runways and aboard aircraft carriers; missiles that exploded on launch pads in Arkansas; and submarines that sank with nuclear weapons aboard in the Atlantic and Pacific.
The Mark 39 bomb that nearly detonated over Goldsboro used a safety architecture called the "ready/safe" system that relied on a specific sequence of mechanical and electrical events occurring in the correct order before the weapon could fire. The fundamental problem was that a bomb falling from a disintegrating aircraft naturally experienced many of those events: the spinning imparted by aerodynamic forces mimicked the spin-motor arming signal, the descent through altitude bands triggered barometric switches, and the deceleration and impact forces could activate acceleration-sensitive mechanisms [2]. The weapon's designers had assumed a controlled release from a stable aircraft — not the chaotic tumbling of a structural breakup.
A 1969 review by Sandia found that in the Goldsboro case, the arm-ready switch in the second bomb — the one that crashed without a parachute — was found in the "armed" position after recovery, meaning that weapon too had come close to completing its firing sequence. Only the physical destruction of the bomb on impact had prevented the electrical circuit from closing [2].
Reform and Legacy
The Goldsboro accident, along with a series of other close calls in the late 1950s and early 1960s, contributed to a complete redesign of nuclear weapon safety systems. The Atomic Energy Commission and the Department of Defense implemented "one-point safe" requirements — ensuring that no single point of failure could lead to nuclear yield — and developed the "strong link/weak link" architecture still used today [4]. Strong links are safety devices designed to survive any accident environment without activating, while weak links are components designed to irreversibly fail (destroying the weapon's firing circuit) in conditions that a strong link might accidentally be defeated — such as fire or extreme impact.
Additionally, coded switch systems (Permissive Action Links, or PALs) were accelerated into deployment, requiring unique codes transmitted from national command authority before a weapon could be armed [4]. These systems directly addressed the Goldsboro scenario — ensuring that environmental forces alone could never replicate the arming sequence.
But the broader lesson of Goldsboro is about institutional secrecy and public trust. For over fifty years, the American public was told this accident was never dangerous — that multiple redundant safety systems made detonation physically impossible. The government knew otherwise. The Jones memo was classified. The one-switch margin was buried in technical reports. The vulnerability of that switch type was documented internally but never disclosed. It took investigative journalism, persistent FOIA requests, and the determination of researchers like Eric Schlosser to reveal that on a January night in 1961, a single low-voltage switch — a component worth a few dollars, with known reliability issues — was all that stood between North Carolina and a thermonuclear catastrophe [3].
The Goldsboro incident remains the closest the United States has come to an accidental nuclear detonation on its own soil. The buried bomb remains in the earth. The secrecy lasted half a century. And the lesson — that complex systems fail in ways their designers never anticipated — remains as relevant as ever.
Sources
- Department of Defense — *Narrative Summaries of Accidents Involving U.S. Nuclear Weapons, 1950–1980*. Link
- Sandia National Laboratories — *Goldsboro Revisited* (Parker F. Jones memo, declassified 2013). Link
- The Guardian — *US nearly detonated atomic bomb over North Carolina* (2013). Link
- National Security Archive, George Washington University — *Nuclear Weapons Accidents*. Link