An analysis of the Hanford data adjusted for age and calendar year of death reduces the number of cancer sites for which a radiation dose relationship can be suggested to two, cancer of pancreas and multiple myeloma. There is no suggestion of a radiation relationship for lymphatic and haemopoietic cancers other than myeloma, or for solid tumors of sites other than the pancreas, with the possible, and very weak, exception of the kidney. The conclusion of Mancuso et al. with regard to variations in sensitivity to radiation by age at exposure appears to be untenable. Radiobiologic considerations, including the results of other studies, suggest that the excess proportional mortality at doses above 10 rem for cancer of pancreas and multiple myeloma is likely to be explainable in terms of a correlate of dose rather than in terms of radiation. The analyses of Mancuso et al. and of the present reviewers are preliminary and do not take advantage of the available data on the total exposed population. A cohort analysis of the Hanford data will permit better understanding of the experience than the present proportional mortality analyses.