High doses of chloroform induced liver cancer in male and female B6C3F1 mice when administered by gavage, kidney cancer in male Osborne-Mendel rats when given by gavage or in the drinking water, and kidney cancer in male BDF1 mice when administered by inhalation. The weight of evidence indicates that chloroform is acting through a nongenotoxic-cytotoxic mode of action. The present study was designed to investigate the dose-response relationships for chloroform-induced lesions and regenerative cell proliferation in B6C3F1 mice as the basis for formulation of a biologically based risk assessment for inhaled chloroform. Different groups of female and male B6C3F1 mice were exposed to atmospheric concentrations of 0, 0.3, 2, 10, 30, and 90 ppm chloroform 6 hr/day, 7 days/week for exposure periods of 4 days or 3, 6, or 13 consecutive weeks. Some additional exposure groups were exposed for 5 days/week for 13 weeks or were exposed for 6 weeks and then examined at 13 weeks. Bromodeoxyuridine was administered via osmotic pumps implanted 3.5 days prior to necropsy, and the labeling index (LI, percentage of nuclei in S-phase) was evaluated immunohistochemically from histological sections. Complete necropsy and microscopic evaluation revealed treatment-induced dose- and time-dependent lesions only in the livers and nasal passage of the female and male mice and in the kidneys of the male mice. Large, sustained increases in the liver LI were seen in the 90-ppm groups at all time points. The female mice were most sensitive, with a no-observed-adverse-effect level (NOAEL) for induced hepatic cell proliferation of 10 ppm. The hepatic LI in the 5 days/week groups were about half of those seen in the 7 days/week groups and had returned to the normal baseline in the 6-week recovery groups. Induced renal histologic changes and regenerative cell proliferation were seen in the male mice at 30 and 90 ppm with 7 days/week exposures and also at 10 ppm with the 5 days/week regimen. Nasal lesions were transient and confined to mice exposed to 10, 30, or 90 ppm for 4 days. In a previous cancer bioassay, a gavage dose of 477 mg/kg/day produced a 95% liver tumor incidence in female B6C3F1 mice. This gavage dose is equivalent to a daily 6 hr/day inhalation exposure of approximately 80 ppm, based on the observed induced increases in the LI as an internal dosimeter. The United States Environmental Protection Agency currently uses the linearized multistage model applied to the mouse liver tumor data from the chloroform gavage study to estimate a virtually safe dose (VSD) as a one in a million increased lifetime risk of cancer. The resulting value is an airborne exposure concentration of 0.000008 ppm. Assuming that chloroform-induced female mouse liver cancer is secondary to events associated with necrosis and regenerative cell proliferation, then no increases in liver cancer in female mice would be predicted at the NOAEL of 10 ppm or below based on the results reported here. Applying an uncertainty factor of 1000 yields an estimate of a VSD at 0.01 ppm. This estimate relies on inhalation data and is more consistent with the mode of action of chloroform.