Mounting epidemiological evidence suggests that smoking may play a role in the etiology of breast cancer. Because smoking-related DNA adducts are detectable in both normal and malignant breast tissues, we hypothesized that breast cancer patients may be sensitive to tobacco-induced carcinogenesis, and this sensitivity could be modulated by variants of metabolic genes. To test this hypothesis, we evaluated benzo(a)pyrene diol-epoxide (BPDE)-induced mutagen sensitivity and polymorphisms of GSTM1 and GSTT1 in a pilot case-control study of breast cancer. Short-term cell cultures were established from blood samples of 100 female breast cancer patients and 105 healthy controls. After 5 h of in vitro exposure to 4 microM of BPDE, we harvested the lymphocytes for cytogenetic evaluation and recorded and compared the frequency of BPDE-induced chromatid breaks between cases and controls. We used a multiplex PCR-based assay to simultaneously detect polymorphisms of GSTM1 and GSTT1 from genomic DNA. We performed univariate and multivariate logistic regression analyses and calculated odds ratios (OR) and 95% confidence intervals (CIs). Cases had a significantly higher frequency of chromatid breaks than did controls (P < 0.0001). The level of chromatid breaks greater than the median value of controls was associated with a >3-fold increased risk of breast cancer [adjusted odds ratio (ORadj) = 3.11; 95% CI = 1.72-5.64]. The risk was more pronounced in those who were < 45 years (ORadj = 4.79; 95% CI = 1.87-12.3), ever-smokers (ORadj = 5.55; 95% CI = 1.85-16.6), alcohol drinkers (ORadj = 4.64; 95% CI = 1.70-12.7), and those who had the GSTT1 null variant (ORadj = 8.01; 95% CI = 1.16-55.3). These data suggest that sensitivity to BPDE-induced chromosomal aberrations may contribute to the risk of developing breast cancer, and such sensitivity may be modulated by both genetic and environmental factors. Larger studies are needed to confirm our findings.