The ability of the rat lymphocyte hprt assay to detect tissue-specific carcinogens was evaluated using 7,12-dimethylbenz[a]anthracene (DMBA) administered under conditions that result in mammary gland tumors. Fifty-day-old female Sprague-Dawley rats were given single doses of 5 and 20 mg/kg DMBA by gavage, and the frequency of 6-thioguanine-resistant (TGr) T-lymphocytes was measured over a period of 21 weeks. A time- and dose-dependent increase in mutant frequency was found, with a maximum frequency found 9-15 weeks after treatment with 20 mg/kg of DMBA. Rats were also dosed with 1, 2.5, 5, 10, 15 and 20 mg/kg of DMBA and assayed for TGr mutant frequency 10 weeks after treatment. A significant linear dose-response was found, with all the DMBA doses resulting in significant increases in mutant frequency. To determine whether or not DMBA-induced mutants in rat lymphocytes reflected the DNA damage in the target tissue, rats were treated with 5 and 20 mg/kg of DMBA and spleen lymphocytes and mammary gland tissue were assayed for DNA adduct formation 1, 3 and 7 days later. A similar pattern of 32P- postlabeled adducts, involving both dG and dA nucleotides, was found in DNA from both the target tissue and the surrogate lymphocytes. Adduct formation was dose responsive in both tissues, with a 2.3- to 4-fold higher concentration in mammary gland as compared with lymphocytes. These results indicate that the rat lymphocyte hprt assay is sensitive to a mammary gland carcinogen and that similar types of DNA adducts are associated with both the lymphocyte mutants and the mammary gland tumors induced by DMBA.