A case-control study of 512 male cases of bladder cancer and 596 male hospital controls (all living in the province of Turin, Northern Italy) has been analyzed for cigarette smoking. Relative risks (RRs) decreased with years since stopping, whereas they seemed not to change with age starting to smoke. A multiplicative effect of RRs for smoking and high-risk occupations was suggested. Filters exerted a protective effect (RR = 0.3 for smokers of filter-tip cigarettes only vs. smokers of non-filter-tip cigarettes). A previously unreported difference was noticed between black types and blond types of tobacco (RR = 0.4 for smokers of the blond types only vs. smokers of the black types, adjusted for age, average daily amount, years since stopping, occupation and use of filter).