The association between alcohol consumption and breast cancer risk was examined in 519 newly incident, histologically confirmed cases of breast cancer diagnosed between 1982 and 1987 within a cohort of 56,837 women enrolled in the Canadian National Breast Screening Study. These women had completed a self-administered food frequency questionnaire including alcohol consumption at enrollment into the study prior to their breast cancer diagnosis. For the total cohort, only a weak association between total alcohol consumption and breast cancer risk is observed, the adjusted relative risk for those drinking 30 or more g/day being 1.22 (95% confidence interval (CI) 0.78-1.90) compared with nondrinkers. There is some evidence for a positive association in women who were premenopausal at the time of enrollment for whom there was a monotonic increase in risk with increasing alcohol intake. Compared with nondrinkers, the adjusted relative risk for alcohol consumption of between 0 and < 10 g of alcohol daily was 1.11 (95% CI 0.71-1.71), between 10 and < 20 g was 1.37 (95% CI 0.79-2.36), between 20 and < 30 g was 1.51 (95% CI 0.80-2.86), and > or = 30 g was 1.86 (95% CI 0.96-3.66; p (trend) = 0.07). These findings contrasted with the results for postmenopausal women where there appeared to be no evidence of any relation. The association in premenopausal women is generally reasonably consistent with that of other studies that have found positive associations with alcohol intake.