Associations between occupation and cancers of the lung (n = 246) and bladder (n = 153) were examined in a case-control study. Controls (n = 212) comprised cases of oral (75%) and pharyngeal cancers (13%) and non-neoplastic oral diseases (12%) at the same hospital. Only males were studied. A personal interview was conducted and a lifetime occupational history and information on demographic and relevant confounding factors including tobacco use were obtained. For lung cases, comparing 'ever' employed with 'never' employed in a particular occupation, significantly elevated risks (adjusted for smoking) were found for textile workers (odds ratio [OR] = 1.99, 95% confidence interval [CI]: 1.3-3.6) and cooks (OR = 4.48, 95% CI: 1.2-16.9). High risks were also observed among ship and dockyard workers (OR = 2.87, 95% CI: 0.8-10.1) and wood workers (OR = 2.88, 95% CI: 0.9-9.6). For bladder cancers, significantly elevated risk was observed only for chemical/pharmaceutical plant workers (OR = 4.48; 95% CI: 1.2-16.5). Two other sets of risk estimates were obtained: one by comparison with a second unexposed group made up of occupations where there was little likelihood of exposure to any cancer-causing occupational agent, and the other by fitting logistic regression models to the data. All methods yielded similar risk estimates. Tobacco smoking but not tobacco chewing was a risk factor for both sites.