To identify occupational cancer risk factors, lifetime occupational, smoking, and alcohol-consumption histories were collected by means of a self-administered questionnaire from 15,463 male cancer patients aged 20 years and over as ascertained from the British Columbia population-based cancer registry; all cases were histologically confirmed. The study methodology, descriptive results, and cancer risks from cigarette smoking are reported. Assessment of questionnaire validation and reliability showed very high correlations between all variables analysed. Non-response bias, assessed among 221 non-responders and 432 matched controls, revealed no statistically significant differences for smoking status, education, or for 11 usual (longest-held job) occupational groups, except for managerial occupations and for four pooled groups that represented 6.7% of all occupations. Except for pancreatic cancer, a significant relationship was found for all cancer sites known to be strongly associated with cigarette smoking.