This epidemiologic evaluation covered autopsies performed at ABCC Nagasaki and at the Nagasaki University School of Medicine, as well as surgical pathology specimens examined at ABCC Nagasaki during 1950-59. The material was related to the ABCC Master Sample population. Bias was demonstrated with respect to symptom status, exposure, age, death certificate diagnosis, male occupations, and history of previous medical examinations at ABCC. In addition the ABCC autopsy series was found to be biased with regard to sex. Comparison with data from Hiroshima indicated similar bias which is possibly more pronounced in Hiroshima. On the basis of these findings, it was concluded that at this time the material is not suitable for broad epidemiologic studies of clinically manifest disease. The material, however, does lend itself to a number of studies of clinically occult disease and other problems which were discussed. 19 references, 32 tables.