Studies conducted in several countries that investigated the relationship of occupation and cancer in men were reviewed and compared. Estimates of the proportion of cancers due to occupational exposure that occurred in the general population were analyzed, and sources of variation were explored. A systematic and standardized evaluation of studies on lung and bladder cancer were undertaken, and only investigations that allowed for confounding from tobacco smoking were included. The proportion of lung cancers attributable to occupation ranged between 1 and 5% (when considering only exposure to asbestos) and 40% (in a study with a high proportion of subjects exposed to ionizing radiation); for bladder cancer, estimates were between 0 and 3% in a few studies and between 16 and 24% in several investigations. No similar attempt of systematic comparison was possible for other cancers.