In 1982, the American Cancer Society enrolled over 1.2 million American men and women in a prospective mortality study of cancer and other causes in relation to different risk factors. The 2-year mortality of 461,981 males aged 40-79 years with known smoking habit has been analyzed in relation to exposure to diesel exhaust (DE) and to employment in selected occupations related to DE exposure. The relative risk (RR) for all causes of death for those exposed was 1.05 (95% confidence interval [CI]: 0.97-1.13). For lung cancer, the RR was 1.18 (95% CI: 0.97-1.44). A dose-response effect was present. Railroad workers, heavy equipment operators, miners, and truck drivers had a higher mortality both for all causes and for lung cancer when compared with subjects with other occupations and no exposure to DE. Truck drivers exposed to DE were not at excess risk of lung cancer if compared with truck drivers unexposed to DE, but a trend of increasing risk with duration of exposure was suggested. DE exposure was also associated with increase in mortality for accidents, cerebrovascular disease, arteriosclerosis, and cirrhosis of the liver. An association based on small numbers was also present for Hodgkin's disease and lymphoid leukemia. No association with chronic non-neoplastic pulmonary diseases or with bladder cancer was found.