We used data from a large population-based case-control study to test the hypothesis that women whose "usual occupation" entailed exposure to higher than background 60-Hz magnetic fields had a higher risk of breast cancer than women without such exposure. Breast cancer cases were identified from four statewide tumor registries, and controls were randomly selected from lists of licensed drivers and Medicare beneficiaries. Information on usual occupation and breast cancer risk factors was obtained by telephone interview. We calculated adjusted odds ratios from logistic regression models for women holding occupations with potential for low, medium, or high magnetic field exposure, compared with background exposure. There was a modest increase in risk for women with potential for high exposure [odds ratio (OR) = 1.43; 95% confidence interval (CI) = 0.99-2.09], and no increase for women with potential for medium (OR = 1.09; 95% CI = 0.83-1.42) or low (OR = 1.02; 95% CI = 0.91-1.15) exposure The risk among premenopausal women in the highest-exposure category was higher (OR = 1.98; 95% CI = 1.04-3.78) than for postmenopausal women (OR = 1.33; 95% CI = 0.82-2.17).