All cause mortality rates for male Oak Ridge workers over the entire follow-up period were similar to national rates (Table I; SMR = 1.00 and 0.96 for white and nonwhite males,respectively). This is unusual in occupational studies where workers in large industries typically show a ''healthy worker effect"--- see Checkoway et al chapter 4. One possible explanation is the large proportion of male workers who were hired at young ages during the war years -- all cause SMR = 1.11  -- and who worked only for a short amount of time. These may have been transient workers not eligible for the draft due to poor health, and/or they may have been subjected to more hazardous working conditions because of the war effort. Notable excesses in mortality occurred only for white males for lung cancer and nonmalignant respiratory disease, and notable deficits occurred for diseases of the digestive system (both malignant and nonmalignant) and diseases of the circulatory system. For both white and nonwhite females the SMRs based on U.S. rates are generally lower than one. This apparent negative bias may be due to underascertainment of vital status. It is known that females are more difficult to follow through some national data sources. If, however, failure to determine vital status is not differentially related to one or more risk factors of interest, then internal comparisons of disease rates are appropriate. This assumption is implicit in our presentation of results for the facility comparison analyses.