First Things First: It Did Happen
Just after midnight on 24 January 1961, a B-52G Stratofortress bomber stationed at Seymour Johnson Air Force Base in Goldsboro, NC, broke up in mid-air and crashed 12 miles north of the base near the cross-roads of Faro, NC.

The aircraft ejected two hydrogen bombs as it fell.

Official Word
Below is the Pentagon's brief narrative of the incident, a copy of which was provided this project by the Stockholm International Peace Research Institute (hereafter SIPRI or Stockholm Institute):

During a B-52 airborne alert mission structural failure of the right wing resulted in two weapons separating from the aircraft during aircraft breakup at 2,000 - 10,000 feet altitude. One bomb parachute deployed and the weapon received little impact damage. The other bomb fell free and broke apart upon impact. No explosion occurred. Five of the eight crew members survived. A portion of one weapon, containing uranium, could not be recovered despite excavation in the waterlogged farmland to a depth of 50 feet. The Air Force subsequently purchased an easement requiring permission for anyone to dig there. There is no detectable radiation and no hazard in the area.
The above narrative, with other official accounts of nuclear-weapons accidents, was entered into the Congressional Record by Louisiana Senator Bennet Johnston on 3 August 1992, and is available by key-word search through the 102nd Congress query page on the Thomas Server.

Some New Information
Chuck Hansen, author of U. S. Nuclear Weapons: The Secret History, and of Swords of Armageddon, a periodical CD-ROM billed as "the world's first updatable nuclear weapons information service," told a reporter for this project that he has examined photographs of the Goldsboro crash site, including pictures of bomb components, and that he has concluded the bombs were, specifically, model MK39 thermonuclear devices.

The MK39, according to Hansen's book cited above, was a cylindrical device measuring 11 feet seven inches in length and weighing between 9,000 and 10,000 lbs. Hansen wrote in 1990 that the potential yield of the MK39s in the Goldsboro incident was "2 to 2.5 megatons" (rather large, as military devices go).

Nuclear Physicist Ralph Lapp caused a stir in 1961 when on page 127 of his just-released book, Kill and Overkill, he said that in the Goldsboro incident the distressed aircraft had jettisoned a "24-megaton bomb." This reference appears to be the origination of what Hansen says is a persistent bit of misinformation on the Goldsboro crash, repeated by Greenpeace, Mother Jones, and most news sources ever since. Hansen says the devices were definitely not 24 megatons in potential yield, but closer to a tenth of that strength. "The United States," Hansen wrote in 1990, "has never deployed such a high-yield weapon" as 24 megatons; and he repeated that assertion in a telephone interview for this project.

The Lapp reference to "24 megatons" may have resulted from a simple decimal deletion.

Hansen told this project that the MK39 may have yielded four megatons at most; and his book reports on ppg. 147-148 that the "largest nuclear weapon ever built by the U.S." was the MARK 17/24, with a potential yield of "15 to 20 megatons."

There is no need to exaggerate the destructive potential of an MK39 device to illustrate a horrible scenario, and 24 megatons is certainly not required. Even four megatons rates as a mighty weapon, more than 250 times the power of the blast that annihilated Hiroshima. According to Dr. Dietrich Schroeer, nuclear physicist and professor of physics at the University of North Carolina at Chapel Hill, the blast from a ground-level detonation of four megatons would have left a crater in the ground a third of a mile wide and leveled homes five miles away, while the heat would have set fires and inflicted third-degree burns to a distance of nine miles from the point of detonation. Many, if not most, U.S. nuclear weapons are smaller in yield. (Today, the premium is placed more upon accurate delivery of devices in the multi-kiloton range. As for the giant MK17/24, it was retired from service in 1958.)

Information taken from the 1987 Nuclear Weapons Data Book Vol. II, published by the Natural Resources Defense Council, lends some support to Hansen's identification of the Goldsboro devices. According to Table 1.2 on page 10 of that publication, approximately 700 MK39s were manufactured, all from 1957 to 1959. According to Figure 1.2 on page 7 of that publication, the MK39 was one of six bomber-deployed nuclear-weapons models in the active stockpile during 1961. No bomber-delivery devices in the 1961 stockpile exceeded 10 megatons in expected yield.

The last of the MK39s was retired from service in 1966.

"Armed" or "Unarmed" Weapons?
Military reports at the time of the accident described the two thermonuclear devices as "unarmed." However, that word is inherently inexact, no matter how it is used. The final "arming" of any military nuclear device requires the completion of numerous steps, executed in the proper sequence and timed correctly. It is thus arguable that any nuclear device could be called technically "unarmed" right up to the moment of its detonation.

Even the account of the accident provided by Hansen sends mixed signals, referring to "unarmed" weapons and "partially armed" weapons, and indicating that at least some of the steps necessary for arming were in fact completed in each of the two bombs. Thus, while the devices may technically have been "unarmed" in that they never detonated, they nonetheless could more accurately have been described throughout the event as "partially armed."

"Unarmed" is a frequently used adjective in military press releases describing broken-arrow incidents. A table beginning on page 65 of SIPRI's 1977 Yearbook presents summary information on 32 such incidents. The arming state of the weapon(s) involved is mentioned in nine of these accounts -- and always the weapons are characterized as "unarmed" or otherwise incapable of nuclear detonation. No weapon is ever described as "armed."

The Goldsboro accident occurred at the height of the Cold War. President John F. Kennedy had taken office only four days earlier, and would soon lead the nation through its closest brush with nuclear war, the Cuban Missile Crisis. Moreover, the B-52 involved in the Goldsboro crash was not on a training flight; it was, according to the Department of Defense account, on an "airborne alert" mission, an operation designed to keep U.S. nuclear arms airborne and deliverable 24 hours a day.

Milton Leitenberg, arms-control specialist and Senior Fellow with the Center for International and Security Studies at the University of Maryland, wrote the SIPRI Yearbook chapter on nuclear-weapons accidents. Leitenberg described to a reporter for this project U. S. airborne-alert activity during the late 50s to early 60s: "In those years, we kept something like 30% of the SAC aircraft in the air at all times, an amazing percentage, and an equal proportion on the runways ready for takeoff at five minutes' notice." He added, "Flights went to turnaround points perhaps two-thirds of the way to their targets. But they were all called training missions, at least, if anything went wrong."

It is predictable and even understandable that military sources would tell the press that the weapons involved were "unarmed." Professor Eric Mlyn of the University of North Carolina at Chapel Hill, who directs The Robertson Scholarship Program and has written extensively on defense policy, says the U.S. and the Soviet Union were in "full-fledged, all-out competition" at the time of the accident and that all matters concerning nuclear weapons were kept very secret. Mlyn said, "Policymakers tried not to talk about nuclear weapons at all," given the controversy they created. [see video]

Given the circumstances and the times, it is unlikely that the Air Force was transporting in its "airborne alert" bomber fleet nuclear weapons which were not fully capable of detonation.

Brush with Catastrophe?
The Stockholm Institute has called the Goldsboro incident "perhaps the single most important example in the published literature of an accident which nearly resulted in a catastrophe." This claim appears to be founded upon yet another hair-raising claim in the 1961 Lapp book.

Lapp wrote in Kill and Overkill that each device involved in the Goldsboro incident was equipped with "six interlocking safety mechanisms, all of which had to be triggered in sequence to explode the bomb." Lapp said that "five of the six interlocks had been set off by the fall..." and thus, "only a single switch prevented the bomb from detonating and spreading fire and destruction over a wide area."

UNC's Schroeer is skeptical that either bomb could have gone off accidentally.
[see video]

However, Lapp's claim that the Goldsboro accident nearly caused a nuclear detonation has found support from some very high sources.

Daniel Ellsberg, famously of the Pentagon Papers case, was quoted in the April 1981 issue of Mother Jones as saying that during his time at the Pentagon he saw "a classified document" about the Goldsboro incident which verified Lapp's claim. Ellsberg stood by his story in a telephone interview for this project, and repeated his 1981 assertions that when the behavior of safety features in both bombs involved in this incident are taken into account, every kind of safety interlock had failed.

Owing to his place in history, Ellsberg is a somewhat controversial figure, but he is very knowledgable about nuclear weapons. Leitenberg said, "However some people may feel about him politically, on this subject Ellsberg is a very credible witness" who was highly placed in the national-security echelons and deeply involved in U.S. weapons-control policy during the years following the Goldsboro accident.

Lapp and Ellsberg have other support in their claims that only a single switch prevented one of the devices from detonating. In September of 1983, Robert McNamara, who served as Secretary of Defense in the Kennedy and Johnson administrations, said at a press conference: "The bombs' arming mechanism had six or seven steps to go through to detonate, and it went through all but one, we discovered later" (Greensboro Daily News, 16 September 1981, pg. A1).

When asked by this project in an email exchange about the story of the failed safety devices, and on which of the two bombs the most of them had failed, Hansen wrote the following on 21 November 2000:

The parachute-retarded weapon came closest to firing; the official reports claim that three of four arming steps were completed. Since the aircraft commander had not thrown the arming switch in the B-52 cockpit, and since that switch was not activated accidentally when the aircraft broke up, it was impossible for the weapon(s) to fire, regardless of how uncomfortably close they came to doing so. This was a very dangerous incident and I suspect that steps were taken afterwards to prevent any repetition of it. I do not now know of any other weapon accident that came this close to a full-scale nuclear detonation (which is not to say that any such incident did not occur later).

Indeed, steps were taken soon after Goldsboro to reduce the threat from accidental detonations. President Kennedy was alarmed by the numbers of nuclear-weapons accidents being brought to his attention. Leitenberg says, "I know that we didn't report them all publicly. We never reported the ones in other countries, or, we only reported the ones that we had to report."

In an effort to reduce the accident threat, the Kennedy administration reduced SAC airborne alert activity. Leitenberg says that the accident rate then "dropped sharply."

President Kennedy also ordered that more elaborate mechanisms be installed on all U.S. nuclear weapons to enhance command and control capability. These peripheral command units are now called permissive action links (PALs) -- an "engineering artifice" which permits detonation only by secret code, and about which Steven M. Bellovin of AT&T Labs Research has prepared an informative page.

Continued Environmental Concerns
One of the two bombs parachuted to earth, imbedding its nose 18" in the ground directly next to Shackleford Road, thus presenting no difficulties for recovery crews. The other bomb hit the ground at high speed with no parachute deployment, disappearing in a farmer's field and leaving only an eight-foot-wide, six-foot-deep crater to indicate where it had entombed itself.

After the accident, the Army Corps of Engineers spent weeks excavating the free-falling bomb's point of impact to a depth of more than 40 feet. At the time, military workers told Gene Price of the Goldsboro News-Argus they were looking for a "lost seat."

The military was never able to recover all of the free-falling bomb's components. The deeper the excavation went, the more problematic soil conditions became. The "lost" bomb fell near Nahunta Swamp, and a high water table characterizes the land in the area. The owner of the field, C. T. Davis, told this project that military workers on the excavation said to him it finally became difficult even to keep the excavating equipment itself from disappearing into the muck.

Rather than continue a losing battle to recover all of the bomb, the military covered over the great hole it had dug, and purchased, for $1,000, an easement from Davis and his heirs. The agreement describes a circle 200 feet in radius where no current or future landowner may dig or drill deeper than five feet, nor ever again use the land "in any manner other than for the growing of crops, the growing of timber, or as a pasture."

(Davis says that two years after the accident, Congress voted him an additional appropriation in settlement for damages and loss of use of the land, but even after all these years Davis refuses to disclose the amount of money he was awarded.)

A woman living in a neat farm house very near the crash scene -- she declined to give her name -- told reporters for this project that she was living in the same house the night of the crash some 40 years ago. "Yes, I was here. It lit up the sky like daylight," she said.

This woman said that while most neighboring homes had been connected to a water main years ago, her home distinctively remains on well water. When asked whether she was aware of any on-going groundwater testing in the area as a result of the crash, she said, "Yes, they still come out and test the water from time to time," but she did not say who "they" were.

According to Dale Dusenbury of the North Carolina Division of Radiation Protection, "they" are the people of his agency. Dusenbury said in a telephone interview that groundwater testing near the Faro crash site is "generally done annually" by the state, but that nothing amiss has been discovered to this point. The tests, according to Dusenbury, have found only "levels of gross alpha, gross beta and total uranium indicative of natural soil."

When asked to speculate on the chances that any radiation would ever leak into the groundwater, Dusenbury said, "The Air Force has been evaluating this, along with other state agencies. If they clean up a place sufficiently to make it environmentally releasable, or if they can establish that what is there is harmless, they will release their ownership. At this time, the Air Force intends to keep its easement because there is still an open question as to whether a hazard exists."

Dusenbury added, "The accident report has never been released to the public or to us. So we intend to keep looking until we know everything is safe."

The last two paragraphs of the Hansen narrative indicate that it is a thermonuclear "secondary" component which remains beneath the earth unrecovered. The secondary forms the "hydrogen" part of a thermonuclear bomb. Secondaries contain not only plutonium and uranium, but a large volume of lithium salt which supplies the hydrogen fuel.

Dusenbury said he does not believe the state has tested the groundwater for lithium. "If it's something essentially inert, it's not a radiological hazard."

Contact us at
Last Updated: 24 January 2001