Ellagic acid, quercetin and robinetin were tested for their ability to antagonize the tumor-initiating activity of benzo[a]pyrene (B[a]P) and (+/-)-7 beta, 8 alpha-dihydroxy-9 alpha, 10 alpha-epoxy-7,8,9,10-tetrahydrobenzo[a]pyrene (B[a]P 7,8-diol-9,10-epoxide-2), the ultimate carcinogenic metabolite of benzo[a]-pyrene. Ellagic acid, robinetin or quercetin (2500 nmol) had no tumor-initiating activity on mouse skin, but the topical application of 2500 nmol of ellagic acid 5 min before a tumor-initiating dose of 200 nmol of B[a]P 7,8-diol-9,10-epoxide-2 caused a 59-66% inhibition in the number of skin tumors per mouse that were observed after 15-20 weeks of promotion with 12-O-tetradecanoylphorbol-13-acetate. Similar treatment with 2500 nmol of robinetin or quercetin caused a statistically insignificant 16-24% inhibition in the tumor-initiating activity of 200 nmol of B[a]P 7,8-diol-9,10-epoxide-2 applied 5 min later. Treatment of mice with 2500 nmol of ellagic acid 5 min before the application of 50 nmol of B[a]P inhibited the mean number of skin tumors per mouse by 28-33% after 15-20 weeks of promotion, but these decreases were not statistically significant. Robinetin and quercetin had little or no effect on the tumor-initiating activity of B[a]P on mouse skin. Treatment of preweanling mice with 1/7, 2/7 and 4/7 of the total dose of ellagic acid (300 nmol), robinetin (1400 nmol), myricetin (1400 nmol) or quercetin (1400 nmol) i.p. on their first, eighth and fifteenth day of life, respectively, did not cause the formation of tumors in animals that were killed 9-11 months later. Similar treatment of preweanling mice with the above doses of the phenolic compounds 10 min before the i.p. injection of a total dose of 30 nmol of B[a]P 7,8-diol-9,10-epoxide-2 during the animal's first 15 days of life caused a 44-75% inhibition in the number of diol-epoxide-induced pulmonary tumors per mouse. Similar treatment with these plant phenols had little or no effect on B[a]P-induced pulmonary tumors.