The purpose of this investigation was to establish a dose response for the effects of dietary phenethyl isothiocyanate (PEITC) on N-nitrosomethylbenzylamine (NMBA)-induced esophageal tumorigenesis and DNA methylation. Groups of 13-27 rats were randomly assigned to AIN-76A diets containing 0, 0.325, 0.75, 1.5 or 3.0 mumol PEITC/g. Two weeks later, rats were administered NMBA subcutaneously at a dose of 0.5 mg/kg once a week for 15 weeks. Animals were maintained on control or experimental diets for an additional 8 weeks and were terminated at week 25 of the experiment. No significant effects on weight gain or food intake were noted for any of the experimental diets when compared with control values. Animals receiving only NMBA developed 9.3 +/- 0.9 tumors/rat, with an incidence of 100%. Dietary PEITC at concentrations of 0.75, 1.5 and 3.0 mumol/g inhibited NMBA-induced esophageal tumor multiplicity by 39%, 90% and 100%, respectively. Esophageal tumor incidence in these groups was reduced by 0%, 40% and 100%, respectively. The 0.325 mumol/g PEITC diet did not significantly affect NMBA-induced esophageal tumorigenesis. These results indicate that the minimum inhibitory dietary concentration of PEITC is between 0.325 and 0.75 mumol/g. Groups of 20 rats were assigned to diets containing 0-3.0 mumol PEITC/g for two weeks as described above, and then sacrificed 24 hours after administration of [3H-methyl]NMBA. The esophageal DNA was isolated, purified, hydrolyzed, and analyzed by HPLC. PEITC inhibited DNA methylation in a dose-dependent manner, as was found in the tumor bioassay. The inhibition of tumor incidence was highly correlated with the percentage inhibition of either 7-methylguanine or O6-methylguanine. These latter results suggest that the inhibitory activity of PEITC in this model is manifested, at least in part, during the functional equivalent of tumor initiation.