The carcinogenicity of 1,4-dioxane was examined by giving groups of 50 F344/DuCrj rats and 50 Crj:BDF(1) mice of each sex 1,4-dioxane in the drinking-water for 2 years. The concentrations of 1,4-dioxane were 0 (control), 200, 1000 and 5000 ppm (wt./wt.) for rats and 0, 500, 2000 and 8000 ppm for mice. The highest dose levels did not exceed the maximum tolerated dose. In the rat, there was significant induction of nasal squamous cell carcinomas in females and hepatocellular adenomas and carcinomas in males and females, peritoneal mesotheliomas in males, and mammary gland adenomas in females. In the mouse, there was significant induction of hepatocellular tumors in males and females. Two nasal tumors occurring in the 8000 ppm-dosed groups were spontaneously rare and, thus, were attributed to 1,4-dioxane exposure. The present studies provided clear evidence of carcinogenicity in rats and mice. Lifetime cancer risk of humans exposed to 1,4-dioxane through drinking-water was quantitatively estimated with a non-threshold approach by application of a linearized multistage model to dose-carcinogenic response relationships, in addition to a threshold approach for estimation of the tolerable daily intake using no-observed- or lowest-observed-adverse-effect levels of the carcinogenic responses and uncertainty factors.