The relationship between smoking and bladder cancer risk was investigated using data from a case-control study conducted between January 1994 and July 1996 in Alexandria, Egypt. Cases were 151 males with incident, histologically confirmed invasive cancer of the bladder, and controls were 157 males admitted to hospital for acute, non-neoplastic, non-urinary tract, non-smoking-related conditions. With reference to never smokers, ex-smokers had a multivariate odds ratio (OR) of 4.4 [95% confidence interval (CI) 1.7-11.7] and current smokers of 6.6 (95% CI 3.1-13.9). The ORs were 5.4 for < 20 and 7.6 for > or = 20 cigarettes per day. After adjustment for cigarette smoking, the ORs were 0.8 for waterpipe and 0.4 for hashish smokers. The risk was significantly related to duration of smoking (OR of 16.5 for > 40 years), and inversely related to age at starting (OR of 8.8 for starting < 20 years), and inversely related to time since quitting smoking. Compared with never smokers who did not report a clinical history of schistosomiasis, the OR was 9.4 for smokers with a history of schistosomiasis, and 10.7 for smokers ever employed in high-risk occupations compared with non-smokers not reporting such a history. Thus, our results, while not giving indications of an increased bladder cancer risk with habits other than cigarette smoking, found a remarkably strong association with various measures of cigarette smoking that could explain 75% of bladder cancer cases among males from Alexandria. The prevalence of smoking was very low among women, and consequently tobacco was not a relevant risk factor for female bladder cancer.