Studies evaluating the association of ovarian cancer with alcohol intake are inconsistent, and few have evaluated this association in the context of folate consumption. Dietary folate and alcohol intakes and lifestyle and medical information were collected with self-administered questionnaires in 1986 from postmenopausal women aged 55-69 followed prospectively for 15 years for risk of epithelial ovarian cancer in the Iowa Women's Health Study. Among 27,205 eligible women free of baseline cancer, 147 incident epithelial ovarian cancer cases were identified by linkage to a cancer registry. Compared to the lowest quartile of total folate (food plus supplement) intake, the multivariable risk ratios (RR) for increasing quartiles were 1.0 (referent), 1.59, 1.24, 1.73 (95% confidence interval [CI], 0.90-3.33; p for trend, 0.20). Compared to non-drinkers, the RRs for increasing alcohol intake were 1.0 (referent), 0.78 for 0.01-3.9 g/d, 0.75 for 4.0-9.9 g/d and 0.58 for > or =10 g/d (95% CI, 0.30-1.11; p for trend, 0.08). Among women with alcohol intake > or =4 g/d compared to < 4 g/d, the apparent risk reduction was limited to those with total folate intake > or =331 microg/d (RR, 0.52; 95% CI, 0.22-1.19; p for interaction, 0.04) although this estimate was based on only seven cases. The association did not change appreciably when we excluded tumors of mucinous histology. These findings suggest that alcohol consumption is inversely related to postmenopausal ovarian cancer, and that the association of folate with ovarian cancer may vary by the amount of alcohol consumed.