Dichloroacetic acid (DCA) and trichloroacetic acid (TCA) are metabolites of the industrial solvent and environmental contaminant trichloroethylene (TCE), as well as contaminants of chlorinated drinking water. Human exposure to these chemicals is of concern as all three have been shown to increase liver tumor incidence in mice. Differences in dose-response curves, progression to cancer, and postexposure regression of lesions suggest that TCA and DCA work through different mechanisms. The purpose of this study was to further characterize the proliferative hepatocellular lesions promoted by TCA and DCA using biomarkers of cell growth, differentiation, and metabolism in liver sections to better delineate the distinctions in the mechanism of the two chloroacetates. Fifteen-day-old female mice were initiated with 25 mg/kg N-methyl-N-nitrosourea. The initiated mice were administered DCA or TCA (20.0 mmol/L) in drinking water from age 49 days until euthanasia at age 413 days. The pathologic assessment showed that the foci of altered hepatocytes and tumors occurring in the animals promoted with DCA were eosinophilic and positive immunohistochemically for TGF-alpha, c-jun, c-myc, CYP 2E1, CYP 4A1, and glutathione S-transferase-pi (GST-pi). The DCA lesions also were essentially negative for c-fos and TGF-beta, but nontumor hepatocytes were consistently TGF-beta-positive. In contrast, tumors promoted by TCA were predominantly basophilic, lacked GST-pi, and stained variably; usually, more than 50% of the tumor hepatocytes were essentially negative for the other biomarkers. This study demonstrates some striking differences in certain molecular biomarkers of cell growth, differentiation, and metabolism between DCA and TCA. The results also suggest some potential growth signal transduction pathways that may contribute to the DCA promotion of tumors, further support the premise that these two chloroacetates promote hepatocarcinogenesis in different ways, and provide a rational basis for a similar comparison with TCE. Such a comparison should give some insight as to whether DCA, TCA, or both are playing a significant role in the murine liver carcinogenesis of the parent compound, TCE.