Bravo Was Supposed to Be 5 Megatons. It Was 15. The Fallout Reached Rongelap and the Record Says They Knew.

Abhishek DashUpdated October 28, 20265 min read
A wooden drafting table lit by an amber drafting lamp, with an oversized hand-drawn mushroom cloud on blueprint paper, a separate sheet showing a wind arrow pointing away from the drawn plume, and a brass compass and red pencil on the desk.

In short

The United States conducted 67 nuclear tests in the Marshall Islands between 1946 and 1958, about 77 megatons in total. Castle Bravo on 1 March 1954 was expected to yield around 5 megatons and produced 15, and the fallout reached Rongelap, Utirik and Rongerik. Eighty-six people on Rongelap and 167 on Utirik were evacuated to Kwajalein within 52 hours. By 1995 the Nuclear Claims Tribunal had awarded 43.2 million dollars, nearly its entire fund.

Key takeaways

  • The United States ran 67 nuclear tests in the Marshall Islands between 1946 and 1958, with a combined yield of roughly 77 megatons.
  • Castle Bravo on 1 March 1954 was predicted at about 5 megatons and produced 15, because lithium-7 in the fuel reacted in a way the designers had not expected.
  • Fallout reached Rongelap, Utirik and Rongerik; 86 people on Rongelap and 167 on Utirik were taken to Kwajalein within 52 hours.
  • Twenty of 29 Rongelap children at the time of Bravo later developed thyroid tumors, according to the record summarised by the Nuclear Claims Tribunal and cited in the historical literature.
  • By 1995 the Nuclear Claims Tribunal had awarded 43.2 million dollars to 1,196 claimants for 1,311 illnesses, nearly its entire fund.

The first thing to get straight about Castle Bravo is that it worked. The device detonated on 1 March 1954 at Bikini Atoll. It produced a fireball, a mushroom cloud, and a crater in the reef. The problem was the size of it. The designers expected about 5 megatons. The device produced 15.

That gap is the whole story. Bravo was the first test in the Castle series, and it was the largest thermonuclear device the United States ever detonated. The extra yield came from lithium-7 in the lithium deuteride fuel, which contributed to the reaction in a way the pre-test models had not captured. The test was a technical success and a public health failure at the same time.

Sixty-seven tests

Between 1946 and 1958, the United States conducted 67 nuclear tests in the Marshall Islands, at Bikini and Enewetak. The combined yield was roughly 77 megatons. That is more than four times the yield of the entire US atmospheric testing program in Nevada.

The tests were not a secret. The Marshallese were moved off Bikini before the first shots in 1946, and the US government told them the relocation was temporary. They were taken to Rongerik, which could not support them. They went hungry. They were moved again, eventually to Kili, a single island with no lagoon and no sheltered fishing grounds.

That sequence is in the congressional record and in the IAEA's later surveys. It is the part of the story that does not depend on the fallout debate. The relocation was badly planned, and the people who were moved paid for it.

The day the wind changed

Bravo was fired at dawn on 1 March 1954. The fallout forecast called for the plume to go one way. The actual wind took it another. The result was that the radioactive coral dust came down on inhabited atolls, including Rongelap, Utirik and Rongerik.

The people on Rongelap saw the snow-like material fall. Children played in it. Within hours, some were nauseous. Within days, their hair began to fall out.

The evacuation came within 52 hours. Eighty-six people on Rongelap and 167 on Utirik were taken to Kwajalein for medical care. That number, 253 people, is in the record. So is the fact that they were evacuated after exposure, not before.

The Japanese fishing vessel Daigo Fukuryu Maru, the Lucky Dragon, was also in the path. All 23 crew members were contaminated. One, Aikichi Kuboyama, died six months later. The incident became a diplomatic problem and a domestic political event in Japan.

What the dosimetry shows

The US later established Project 4.1 to study the exposed Marshallese. The records show a range of acute effects and long-term thyroid outcomes. Fifteen islands and atolls were contaminated. Twenty of the 29 Rongelap children at the time of Bravo later developed thyroid tumors, according to the account summarised in the historical literature and the tribunal record.

Those numbers are the core of the documented case. They are also where careful readers should slow down. The dose reconstruction is contested, the medical follow-up is incomplete, and the literature contains both strong findings and methodological arguments about them. The thyroid tumors are the best-documented outcome. The broader claim that every health problem in the islands traces to the tests is not something the record supports.

The compensation fund

The Compact of Free Association, signed in 1986, created the Nuclear Claims Tribunal and gave it a fund. Section 177 of the Compact set out the claims process. The tribunal was not a cleanup agency. It paid people for illnesses and property loss.

By 1995, according to the tribunal's own reporting, it had awarded 43.2 million dollars to 1,196 claimants for 1,311 illnesses. That was nearly its entire fund. Later claims ran into a fund that was already close to empty.

This is the mechanism that matters. The United States did not rebuild the atolls. It created a claims process, funded it once, and left the rest to a tribunal that could award money it did not have. The paper trail here is a budget line, not a conspiracy.

What the record does not say

The popular version of this story often says the government knew the atolls would be hit and tested anyway. The documents support a narrower claim. The test had a fallout forecast. The forecast was wrong. The atolls were inhabited and downwind. The evacuation came after the exposure.

That is enough for accountability without inventing intent. The miscalculation was technical. The decision to keep testing in the same region, with the same kind of forecast, is a policy choice that the record shows was made repeatedly.

The second limit is the medical record. The thyroid findings are strong. The claim that radiation caused every later cancer in the islands is not established, and the studies that try to answer it are limited by small populations, incomplete dosimetry, and decades of missing follow-up.

The third limit is the archive itself. Some of the test data is still classified. The IAEA surveys and the DOE records are the best public sources, and they do not answer every question about individual exposure.

Why this belongs in the record

The Marshall Islands case is often filed under history, which makes it easy to treat as settled. It is not. The compensation fund was exhausted. The atolls were not fully remediated. The people who were moved are still dealing with the consequences of a relocation that was supposed to be temporary.

The documents are public. The IAEA surveyed Bikini. The tribunal published its awards. The DOE records the tests. What is missing is not evidence that something happened. It is a complete accounting of what it cost, and who paid.

On this page

Sources

  1. Radiological Conditions at Bikini Atoll: Prospects for Resettlement (IAEA report)International Atomic Energy Agency, 1998
  2. Castle Bravo, 1 March 1954: test record and fallout reconstructionUS Department of Energy, 1954
  3. Nuclear Claims Tribunal, Republic of the Marshall IslandsNuclear Claims Tribunal, 1995
  4. Compact of Free Association, Title II, Section 177US Department of the Interior, 1986
  5. Castle Bravo: the yield, the fallout and the evacuationsAtomic Heritage Foundation, 2016

Frequently asked questions

Was Bravo an accident?

The yield was a miscalculation, not an explosion that went out of control. The designers expected roughly 5 megatons. Lithium-7 in the lithium deuteride fuel contributed to the reaction in a way the pre-test models did not capture, and the device produced 15 megatons. The test worked. The fallout forecast did not.

Did the United States know people would be exposed?

The record shows the test was planned with a fallout forecast and that the forecast was wrong. It also shows that inhabited atolls sat downwind and that the evacuation came after the fact. Whether the planners believed the atolls were safe is a question the documents answer only in part; the documents do show that the risk was understood in general terms and that the actual dose was not.

What did the Marshallese get?

A compensation system, not a cleanup. The Compact of Free Association created the Nuclear Claims Tribunal in 1986. By 1995 the tribunal had awarded 43.2 million dollars to 1,196 claimants for 1,311 illnesses, nearly its entire fund. Later claims ran into a fund that was already close to empty, and the contaminated atolls were not made habitable again.