Chlorination of water rich in organic material is known to produce a complex mixture of organochlorine compounds, including mutagenic and carcinogenic substances. A historical cohort study of 621,431 persons living in 56 towns in Finland was conducted in order to assess the relation between historical exposure to drinking water mutagenicity and cancer. Exposure to quantity of mutagenicity was calculated on the basis of historical information of raw water quality and water treatment practices using an empirical equation relating mutagenicity and raw water pH, KMnO4 value and chlorine dose. Cancer cases were derived from the population-based Finnish Cancer Registry and follow-up time in the study started in 1970. Age, gender, time period, social class, and urban residence were taken into account in Poisson regression analysis of the observed numbers of cases using expected numbers of cases standardized for age and gender as a basis. Excess risks were calculated using a continuous variable for mutagenicity for 3,000 net rev/l exposure representing an average exposure in a town using chlorinated surface water. After adjustment for confounding, a statistically significant excess risk was observed for women in cancers of the bladder (relative risk [RR] = 1.48, 95 percent confidence interval [CI] = 1.01-2.18), rectum (RR = 1.38, CI = 1.03-1.85), esophagus (RR = 1.90, CI = 1.02-3.52), and breast (RR = 1.11, CI = 1.01-1.22). These results support the magnitude of excess risks for rectal and bladder cancers found in earlier epidemiologic studies on chlorination by-products and give additional information on exposure-response concerning the mutagenic compounds. Nevertheless, due to the public health importance of water chlorination, uncertainty related to the magnitude of observed risks, and the fact that excess risks were observed only for women, the results of the study should be interpreted with caution.