The proportions of male lung cancers due to occupational exposure and, respectively, to cigarette smoking in a highly industrialized area of Northern Italy were estimated in a population-based case-control study in 1976-9. Two hundred and four out of the 211 lung cancer cases and 351 controls sampled from the source population were questioned about their occupational and smoking histories. On the basis of the occupational history each subject was classified as probably exposed (+), possibly exposed (?), or unexposed (-) to one or more of the chemicals known to be carcinogenic for the human lung, namely asbestos, polycyclic aromatic hydrocarbons, arsenic, nickel and chromium compounds, BCME, CMME and vinyl chloride. Upon stratification by cigarette smoking, contrasting the occupationally exposed subjects, whether certainly or uncertainly defined, with the unexposed ones, the RR for lung cancer was 2.1 and the occupational etiologic fraction was 0.33 (95% confidence interval 0.19-0.47). The tobacco etiologic fraction was 0.81, while the two exposures together accounted for 89% of the total burden of incident cases. If 33% of all male lung cancers were of occupational aetiology, then this alone would represent 5% of all cancer deaths.