A cohort study was done on 1396 deaths seen among 4352 Japanese male Zen priests during a follow up period from 1 January 1955 to 31 December 1978. Standardised mortality ratios were computed for major causes of death by comparing with the counterparts of the general Japanese male population. The SMR for all causes of death was 0.82 (p less than 0.001) and the SMR values for cerebrovascular diseases, pneumonia and bronchitis, peptic ulcer, liver cirrhosis, cancer of the respiratory organs, and cancer of the lung were all significantly smaller than unity. Taking regional mortality differences into account, a similar computation was made dividing the cohort into two subcohorts--that is, the priests living in eastern Japan and those in western Japan. Both subcohorts showed a highly significantly smaller SMR than unity for all causes of death. With the exception of only a few causes of death for which the observed number of deaths was small, they also showed such reduced SMRs for nearly all of the causes of death tested. A questionnaire survey on the current life style of active priests showed that they smoke less, eat less, meat and fish as they follow the more traditional Japanese dietary habits, and live in less polluted areas, but their drinking habits do not differ much from that of the average Japanese adult man. Possible reasons for their reduced mortality are discussed.