A population-based case-control study was conducted in Washington County, Maryland (United States) to explore the association between incident bladder cancer and exposure to drinking water from chlorinated surface sources. Cancer cases were White residents, enumerated in a 1975 county census and reported to the Washington County Cancer Registry (n = 294) between 1975 and 1992. White controls, frequency matched by age (+/- 5 years) and gender, were selected randomly from the census (n = 2,326). Households receiving municipal water, which generally derived from chlorinated surface waters, were treated as having 'high exposure' and all others, as 'low exposure.' Duration of exposure to type of drinking water was based on length of residence in the census household prior to 1975. Odds ratios (OR) were calculated using logistic regression methods, adjusting for age, gender, tobacco use, and urbanicity. Bladder cancer risk was associated weakly in the general population with duration of exposure to municipal water. The association was limited to those who had smoked cigarettes. In ever-smokers compared with never-smokers with low exposure, the adjusted ORs for bladder cancer risk with increasing exposure were 1.3, 1.4, 1.4, 1.7, 2.2, 2.8, respectively, for 0, 1-10, 11-20, 21-30, 31-40, > 40 years' exposure duration. The ORs in smokers were not diminished after adjusting for smoking history and intensity.