The induction of hepatic tumors in B6C3F1 mice treated with trichloroethylene (TRI) has been attributed to its metabolism to trichloroacetate (TCA). Trichloroacetate is an effective peroxisome proliferator in mice at blood concentrations that are readily achieved with carcinogenic doses of TRI. Recent data has demonstrated that both TCA and dichloroacetate (DCA) are capable of inducing liver tumors in B6C3F1 mice. Although long recognized as a metabolite of TRI, little attention has focussed on the role DCA might play in the hepatocarcinogenic effects of TRI. There are significant differences in the effects of DCA and TCA on the liver of B6C3F1 mice. Trichloroacetate treatment induces peroxisome proliferation, increases lipid deposition, and results in a marked accumulation of lipofuscin in the liver with long-term exposures. Dichloroacetate induces a markedly enlarged liver associated with a cytomegaly and large accumulations of glycogen. The cytomegaly is associated with the development of focal areas of recurrent liver necrosis which in tum lead to high levels of cell proliferation in the area surrounding these lesions. Induction of peroxisomes with DCA is transitory and the accumulation of lipofuscin is much less evident than with TCA treatment. Studies of TRI metabolism demonstrate that blood levels of DCA produced are sufficient to account for the hepatocarcinogenic effects of TRI. The rather low concentrations of DCA found in the urine of mice treated with TRI relative to TCA concentrations are due to the much more rapid and complete metabolism of DCA. These data do not support the conclusion that the hepatocarcinogenic effects of TRI are simply related to peroxisome proliferation.