Of the estimated 205,000 military personnel who participated in the US atmospheric nuclear weapons testing program from 1945 to 1962, less than 1% had ionizing radiation doses that met or exceeded the current federal occupational guideline for dose of 5 rem (roentgen equivalents in humans) in a 12-month period. The objective of this study was to determine whether veterans who received the highest gamma radiation doses (n= 1010) have experienced increased cancer mortality compared with a group of Navy veterans who received a minimal radiation dose as participants of HARDTACK I (n= 2870). Mortality from all causes of death (relative risk, 1.22; 95% confidence interval, 1.04 to 1.44) and from all lymphopoietic cancers (relative risk, 3.72; 95% confidence interval, 1.28 to 10.83) was significantly elevated among the 5-rem cohort compared with the Navy controls. The lack of statistically significant excesses in deaths from many of the known radiogenic cancers suggests that the observed excess mortality may be the result of many factors, of which radiation exposure was only one.