leukemia were not a chance occurrence, the SMOKY participants may have received higher radiation doses than supposed (perhaps from neutrons or inhaled radioactive material not detected by film badges) or radiation was more carcinogenic at low doses than previously assumed. The CDC published a preliminary report of these findings in the 3 October 1980 edition of the Journal of the American Medical Association (4). The CDC continued to study the incidence of all forms of cancer as well as causes of death among the cohort, which was eventually refined to 3,217 veterans. Disease incidence and mortality data were collected through 1979 on over 95 percent of the cohort. The followup study identified a total of 112 cancer cases, which is below the expected number of 117.5 cases. The incidence of some specific cancer types was slightly higher than expected, but the increase was not considered statistically significant with the exception of leukemia (one additional case was identified). Cancers of the digestive system, respiratory, genital, and urinary systems occurred less often than expected. No cancers of the bone/ joints, soft tissue, endocrine system, or multiple myeloma were found. With regard to mortality, the cohort had considerably fewer total deaths than expected. The number of deaths increased in only three categories-- infectious and parasitic diseases, accidents, and killed in action. Deaths from individual types of cancer exceeded the norm in five categories-leukemia, brain and nervous system, eye and orbit, genital system, and skin melanoma. Again, only the increased incidence of leukemia deaths was found statistically significant. An analysis of the film badge dosimetry available for the cohort showed that, in general, radiation doses were well within current occupational exposure standards. The analysis also showed that the mean dose received by participants engaged in the military maneuver was higher than the mean dose received by support units. However, the frequency of cancer was higher among the participants assigned to support units. Assuming that the dosimetry is correct, at least in a relative sense, the opposite would be expected if radiation were the cause. 183