The carcinogenic activity of chloroform administered at 0, 200, 400, 900, and 1800 mg/liter in drinking water was studied in male Osborne-Mendel rats and female B6C3F1 mice. A second control group was included in the study and was restricted to the water consumption of the high-dose group. Animals were maintained on study for 104 weeks. Group sizes were adjusted at low doses such that a detectable tumor response would result at the lowest dose if there was a linear relationship with dose, and the higher doses produced responses similar to previous carcinogenesis bioassays of chloroform. The primary finding was that chloroform increased the yield of renal tubular adenomas and adenocarcinomas in male rats in a dose-related manner. For the high-dose group, which corresponded to a time-weighted average dose of 160 mg/kg per day for 104 weeks, there was a 14% incidence of renal tubular adenomas and adenocarcinomas, vs 1% in the control group. This compares to a 24% incidence observed when 180 mg/kg per day of chloroform was administered for 78 weeks in earlier studies. In contrast, chloroform in the drinking water of mice failed to increase the incidence of hepatocellular carcinomas in female B6C3F1 mice. The highest dose group received a time-weighted average dose of 263 mg/kg for 104 weeks, resulting in a 5% combined incidence of hepatocellular adenoma and carcinoma relative to a 6% incidence in the control groups.(ABSTRACT TRUNCATED AT 250 WORDS)