The effect of voluntary exercise on azoxymethane-induced hepatocarcinogenesis was investigated in male F344 rats. Beginning at 5 weeks of age, all animals were divided into two groups (sedentary and exercise) and fed AIN-76A semipurified diet ad libitum. At 7 weeks of age, animals were given azoxymethane (AOM) s.c. at a dose level of 15 mg/kg of body weight, once weekly for 2 weeks. Four days after the second dose of AOM, all animals in the exercise group were housed in individual wheel-cage units and the animals in the sedentary group were housed in plastic cages. The experiment was terminated at 38 weeks post-AOM treatment. Body weights of animals in the exercise and sedentary groups were comparable. Immunohistochemical staining of glutathione S-transferase placental form (GST-P) was performed in the liver and measured GST-P positive foci. Density (number of GST-P positive foci/cm2 area of liver section), average area of foci and unit area of foci were significantly inhibited in the exercise group, although the incidence of neoplastic nodules and GST-P positive foci were unaffected by the exercise. Thus, energy expenditure due to exercise may reduce hepatocarcinogenesis in a laboratory animal model.