Aberrant crypt foci (ACF) consisting of one more single aberrant crypts (AC) are putative preneoplastic lesions that have been proposed as intermediate biomarkers for colon cancer. Using ACF as the end-point we have studied the effects of two different classes of colon carcinogens, 1,2-dimethylhydrazine dihydrochloride (DMH; 10 or 20 mg/kg body wt/injection) or its metabolite azoxymethane (AOM; 5 mg/kg) and 3,2'dimethyl-4- aminobiphenyl hydrochloride (DMAB; 50 mg/kg) in F344 and Lewis rats. Each carcinogen was given alone or DMH/AOM and DMAB were given in combination in either alternating or successive order in multiple doses. Each compound given alone induced ACF in both rat strains and the effect was most pronounced in the F344 rats. DMAB, not previously tested for ability to induce ACF in rats, was clearly less potent than DMH or AOM. The highest number of ACF was found distally in the colon, independent of treatment or rat strain. Surprisingly, DMAB markedly decreased the carcinogenic effect of DMH, evaluated both as numbers of ACF and AC per colon, as well as number of ACF with four or more AC, when both classes of carcinogens were given alternately. A more pronounced reduction was found in F344 rats than in Lewis rats, being 75-77% and 64-68% respectively with the highest DMH dose. The same tendency was found with successive exposure to DMAB followed by DMH or AOM. These differences in timing of exposure and the different metabolic pathways used by the two classes of carcinogens make a metabolic interaction unlikely as the reason for the antagonistic effect of DMAB on DMH or AOM. The type of standard diet used was found to influence the induction of ACF by the colon carcinogen DMH.