A major goal in molecular epidemiology is to identify preventable environmental risk factors and susceptible subpopulations. In a hospital-based molecular epidemiological case-control study of breast cancer, we investigated the relationship between DNA damage from exposure to polycyclic aromatic hydrocarbons (PAHs) and susceptibility attributable to inherited deletion of the xenobiotic detoxifying gene, glutathione S-transferase M1 (GSTM1). Prior to breast surgery, women (n = 227) were enrolled and interviewed and donated a blood sample. PAH-DNA adduct levels were measured by immunohistochemistry in breast tissue samples retrieved from pathology blocks, and GSTM1 genotype was determined by PCR using WBC DNA. The GSTM1 analysis included 95 cases and 87 benign breast disease controls. GSTM1 genotype was not associated with breast cancer case-control status (odds ratio = 0.73; 95% confidence interval, 0.37-1.44). However, the GSTM1 null genotype predicted PAH-DNA adduct levels in malignant (beta = 0.407; P = 0.003) and nonmalignant (beta = 0.243; P = 0.05) breast tissue from cases. This relationship was not seen in tissue from controls (beta = 0.095; P = 0.341). When tissue from controls was compared with tumor tissue from cases, there was a significant case-control difference in PAH-DNA adduct levels among women who were GSTM1 null. There was no such case-control difference among women who were homozygous or heterozygous for GSTM1. There was an interaction between GSTM1 and case-control status on adduct levels in breast tissue (P = 0.002). The results suggest that genetic susceptibility to the formation of PAH-DNA adducts in breast tissue may play a role in breast cancer development.