Although the results of cohort studies on the association between alcohol and breast cancer are rather consistent, the current evidence is based solely on North American cohorts. Therefore, this association was evaluated in the Netherlands Cohort Study on diet and cancer, conducted since 1986 among 62,573 women aged 55-69 years. After 3.3 years of follow-up (1986-1989), 422 incident breast cancer cases for which there were complete alcohol consumption data were available for analysis. In multivariate case-cohort analyses, the rate ratio for breast cancer in drinkers versus nondrinkers was 1.31 (95 percent confidence interval 1.00-1.71). When separate alcohol intake categories were compared with nondrinking, the rate ratios were 1.30, 1.29, 1.28, and 1.72 for women who consumed < 5, 5-14, 15-29, and > or = 30 g of alcohol per day, respectively (trend p = 0.047). Whereas beer consumption was not associated with breast cancer risk, increased risks were found at higher levels of both wine and liquor consumption. The alcohol-breast cancer association was found to be stronger among women with a history of benign breast disease, women with a history of breast cancer among sister(s), and women with an early menopause, and it varied considerably according to age at first birth. These results support a positive association between alcohol and breast cancer among postmenopausal women. The increased risk was particularly found among women who consumed 30 g or more of alcohol daily.