To identify possible extrinsic and intrinsic DNA-damaging factors involved in breast cancer etiology, we measured the level of aromatic and lipid peroxidation-related DNA adducts in samples of normal tissue adjacent to breast tumors obtained from 87 breast cancer patients using 32P postlabeling. Twenty-nine cancer-free women who underwent reduction mammoplasty served as controls. Tissue samples from the breast cancer patients contained significantly higher levels of aromatic DNA adducts (mean +/- SEM: 97.4 +/- 23.4 x 109 nucleotides) than did samples obtained from the controls (mean +/- SEM: 23.5 +/- 6.9 x 109 nucleotides). A bulky benzo[a]pyrene (BP)-like adduct was detected in 41% of the cancer patients, but in none of the controls. The level of this adduct was extremely high in some patients (> 1/106). While 88% of the patients with a smoking history had smoking-specific DNA adducts in their breast tissues, the presence of BP-like adduct was not related to smoking history. The cancer patients also had a significantly higher level of lipid peroxidation-related DNA adducts than did controls. The level of these adducts correlated with the presence of the BP-like adduct. To further explore the origin of the BP-like adduct, we examined the level of polycyclic aromatic hydrocarbon (PAH)-DNA adducts and 8-hydroxyguanine (8-OH-G) in tissue sections obtained from 37 breast cancer patients using immunocytochemistry. We found that patients who had the BP-like adduct showed significantly greater immunostaining for PAH adducts than did those without the BP-like adduct (p = 0.04). In addition, we found that adipocytes tended to have greater immunostaining for the PAH adducts than did epithelial cells. On the other hand, epithelial cells tended to have a higher frequency and greater intensity of staining for 8-OH-G than did adipocytes. The detection of PAH adducts, lipid peroxidation-related DNA adducts, and 8-OH-G in normal breast tissues of breast cancer patients suggests that both exogenous and endogenous DNA-damaging factors may be involved in breast cancer. The exogenous source may involve the types of carcinogen exposure other than cigarette smoking, and the endogenous source may involve oxidative stress associated with normal metabolic activities.