We examined the role of farming and pesticide exposure among 862 cases and 790 controls in a population-based, case-control study of breast cancer conducted in North Carolina between 1993 and 1996. We obtained exposure information through personal interview. Increasing duration of farming was inversely associated with breast cancer risk; odds ratios (95% confidence intervals) were 1.2 (0.8-1.7), 0.8 (0.5-1.2), 0.7 (0.5-1.1), and 0.6 (0.4-0.9) for 1-10, 11-17, 18-23, and >23 years of farming, respectively, relative to nonfarmers. Inverse associations persisted when farming was restricted to calendar time periods of 2,2-bis(p-chlorophenyl)- 1,1,1-trichloroethane (DDT) use or to farming at ages 9-16. Among women who farmed, odds ratios (ORs) were elevated for those who reported being present in fields during or shortly after pesticide application (OR = 1.8, 95% CI = 1.1-2.8) and for those who reported not using protective clothing while applying pesticides (OR = 2.0; 95% CI = 1.0-4.3), but not among those who reported using protective clothing (OR = 0.8; 95% CI = 0.4-1.8). We conclude that residence or work on farms may be associated with a reduced risk of breast cancer. Nevertheless, our results suggest a possible increased risk of breast cancer among a subgroup of farming women who were most likely to be exposed to pesticides.