Acetaminophen (ACT), the most commonly used analgesic and antipyretic in the United States, was previously demonstrated to be a hepatocarcinogen in one mouse study but not in rats. In order to help elucidate the potential mechanisms of carcinogenesis by this nongenotoxic chemical and its relationship to hepatotoxicity, ACT was fed to groups of 60-120 male B6C3F1 mice at dietary concentrations of 5000 or 10,000 ppm from 6 weeks of age for periods of up to 70 weeks to study the hepatotoxic effects of ACT. To test for potential liver tumor-promoting effects of ACT, N-nitrosodiethylamine (DEN) was injected intraperitoneally at 40 mg/kg into additional groups of 30-60 male B6C3F1 mice at 4 weeks of age. Two weeks later some mice received ACT at dietary concentrations of 5000 or 10,000 ppm. Mice were sacrificed either at 24 weeks after DEN injection or after 22 or 70 weeks of ACT exposure. The livers were weighed and prepared for qualitative and quantitative histological evaluation of focal hepatocellular proliferative lesions (FHPL) including microscopic hyperplastic foci and neoplasms by automated image analysis. At 24 weeks the incidence and number of FHPL per square centimeter were significantly increased only in DEN-treated mice receiving 10,000 ppm ACT. Chronic hepatotoxicity was mild at this time. At 72 weeks ACT alone had no effect on the incidence or number of naturally occurring liver tumors despite severe chronic hepatotoxicity and suppression of body weight gain in mice receiving 10,000 ppm and only mild toxicity at 5000 ppm. There were histological findings suggesting that the chronic hepatotoxicity had, in part, a vascular pathogenesis. This study provided evidence against the hypothesis that chronic hepatotoxicity, in and of itself, results in an increased incidence of naturally occurring liver tumors in mice.