Consumers of chlorinated drinking water have a small excess of bladder cancer. Risk may have been underestimated because of confounding and misclassification of exposure status. To address these problems, we undertook a case-control study. Detailed residential histories were obtained by telephone interviews with informants of 614 individuals who died of primarily bladder cancer and 1,074 individuals who died of other causes. Their surface water has been disinfected with chlorine or a combination of chlorine and ammonia (cloramine) since 1938. The mortality ratio for bladder cancer among individuals who resided only in communities supplied with drinking water disinfected with chlorine, relative to individuals who resided only in communities supplied with drinking water disinfected with chloramine, was 1.6 (95% confidence interval = 1.2-2.1), using all controls; when the comparison group was restricted to individuals who died of lymphoma, the mortality odds ratio was 2.7 (95% confidence interval = 1.7-4.3).