Previous studies reported an association between leukemia rates and amounts of fallout in southwestern Utah from nuclear tests (1952 to 1958), but individual radiation exposures were unavailable. Therefore, a case-control study with 1177 individuals who died of leukemia and 5330 other deaths (controls) was conducted using estimates of dose to bone marrow computed from fallout deposition rates and subjects' residence locations. A weak association between bone marrow dose and all types of leukemia, all ages, and all time periods after exposure was found. This overall trend was not statistically significant, but significant trends in excess risk were found in subgroups defined by cell type, age, and time after exposure. The greatest excess risk was found in those individuals in the high-dose group with acute leukemia who were younger than 20 years at exposure and who died before 1964. These results are consistent with previous studies and with risk estimates for other populations exposed to radiation.