Mortality from several cancers, including bladder cancer, is elevated in a Taiwanese population exposed to high levels of arsenic in drinking water. Data from the Utah respondents to the National Bladder Cancer Study conducted in 1978 were used to evaluate these associations in a US population exposed to measurable, but much lower, levels of drinking water arsenic. Two indices of cumulative arsenic exposure were used, one representing total cumulative exposure (index 1) and the other, intake concentration (index 2). Overall, there was no association of bladder cancer with either measure; however, among smokers, but not among nonsmokers, positive trends in risk were found for exposures estimated for decade-long time periods, especially in the 30- to 39-year period prior to diagnosis. Exposures were in the range 0.5-160 micrograms/liter (mean, 5.0 micrograms/liter). The data raise the possibility that smoking potentiates the effect of arsenic on risk of bladder cancer. However, the risk estimates obtained are much higher than predicted on the basis of the results of the Taiwanese studies, raising concerns about bias or the role of chance. Confirmatory studies are needed.