The etiology of human breast cancer is currently undefined. However, it has been hypothesized that exposure to chemical carcinogens may be an important factor. Extrapolation from rodent models for chemically-induced mammary cancer suggests the possibility that human mammary epithelial cells in situ might contain DNA adducts due to exposure to environmental chemicals. We have therefore screened breast epithelial cells from 10 donors for the existence of DNA adducts using the 32P-postlabeling assay. In order to validate this analysis technique, we also examined the DNA adducts formed in human mammary cells exposed to benzo[a]pyrene (B[a]P) in vitro, and adducts formed in rat mammary epithelial cells exposed to B[a]P in vitro and in vivo. Confirming previous results using HPLC analysis of [3H]B[a]P-DNA adducts, the major B[a]P-DNA adduct formed by human mammary epithelial cells in vitro was (+)-anti-B[a]P-7,8-dihydrodiol-9,10-epoxide (BPDE):deoxyguanosine. This adduct did not appear to be formed by rat mammary cells exposed to B[a]P in vitro. However, 32P-postlabeling analysis of mammary epithelial cell DNA from rats exposed to B[a]P in vivo indicated that (+)-anti-BPDE-deoxyguanosine was a major B[a]P-DNA adduct under these exposure conditions. When the mammary epithelial cells from 10 human donors were screened for DNA adducts formed in situ, cells from three donors exhibited distinct adduct patterns. None of these adducts appeared to be (+)-anti-BPDE-deoxyguanosine. The existence of DNA adducts in human mammary epithelial cells in situ, coupled with the data indicating that rat mammary cells form different B[a]P adducts in vitro and in situ, suggests the need for further study of human breast cell adducts.