Facts concerning a NIOSH study of occupational exposure to low level radiation among workers at the Portsmouth Naval Shipyard are reviewed. The contention that this study contained a number of serious analytic mistakes is challenged. The history of this mortality study of 27,000 current and former workers is briefly recounted. For each of 83 specific causes of death, NIOSH researchers calculated expected numbers of deaths adjusted for age, sex, race, and calendar time period and stratified by 5 year latency periods, and (for radiation workers) by cumulative radiation exposure. The study produced no evidence to support the previously reported 5 fold excess in leukemia mortality nor any statistically significant excesses in mortality from any other causes of death examined. Interpretation of the data produced during the study is described. It is suggested that previous researchers, apparently in the absence of any initial hypotheses, examined the NIOSH data for possible associations between radiation exposure and mortality from any cause. They reported a positive relationship between radiation exposure and lung cancer in workers with more than 15 years of latency and 1 rem cumulative radiation exposure. The author concludes that while this finding deserves further examination, conclusions were not scientifically drawn. Overinterpretation of their findings illustrates the pitfalls of evaluating complex data without benefit of an initial hypothesis.