Several investigators have suggested that certain hydroxylated metabolites of 17 -estradiol (E 2 ) are the proximate carcinogens that induce mammary carcinomas in estrogensensitive rodent models. The studies reported here were designed to examine the carcinogenic potential of different levels of E 2 and the effects of genotoxic metabolites of E 2 in an in vivo model sensitive to E 2 -induced mammary cancer. The potential induction of mammary tumors was determined in female ACI rats subcutaneously implanted with cholesterol pellets containing E 2 (1, 2, or 3 mg), or 2-hydroxyestradiol (2-OH E 2 ), 4-hydroxyestradiol (4-OH E 2 ), 16 -hydroxyestradiol (16 -OH E 2 ), or 4-hydoxyestrone (4-OH E 1 ) (equimolar to 2 mg E 2 ). Treatment with 1, 2, or 3 mg E 2 resulted in the first appearance of a mammary tumor between 12 and 17 weeks, and a 50% incidence of mammary tumors was observed at 36, 19, and 18 weeks respectively. The final cumulative mammary tumor incidence in rats treated with 1, 2, or 3 mg E 2 for 36 weeks was 50%, 73%, and 100% respectively. Treatment of rats with pellets containing 2-OH E 2 , 4-OH E 2 , 16 -OH E 2 , or 4-OH E 1 did not induce any detectable mammary tumors. The serum levels of E 2 in rats treated with a 1 or 3 mg E 2 pellet for 12 weeks was increased 2-to 6-fold above control values (30 pg/ml). Treatment of rats with E 2 enhanced the hepatic microsomal metabolism of E 2 to E 1 , but did not influence the 2-or 4-hydroxylation of E 2 . In summary, we observed a dosedependent induction of mammary tumors in female ACI rats treated continuously with E 2 ; however, under these conditions 2-OH E 2 , 4-OH E 2 , 16 -OH E 2 , and 4-OH E 1 were inactive in inducing mammary tumors.