The relation between alcohol consumption and the risk of development of hypertension was studied among 58,218 US female registered nurses aged 39 to 59 years who were free of diagnosed systemic hypertension and other major diseases. In 1980, all of these women completed an independently validated dietary questionnaire, which included use of alcoholic beverages. During 4 years of follow-up, 3,275 women reported an initial diagnosis of hypertension; validity of the self-report measure was demonstrated in a subsample. When compared to nondrinkers, women drinking 20 to 34 g of alcohol per day (about 2 or 3 drinks) had a significantly elevated relative risk of 1.4; the 95% confidence interval (CI) was 1.2 to 1.7 after adjustment for age and Quetelet's index. For women consuming greater than 35 g/day, the relative risk was 1.9 (95% CI 1.6 to 2.2). Adjustment for smoking and dietary variables did not alter these results. Independent significant associations were observed for the consumption of beer, wine and liquor. These prospective data suggest that alcohol intake of up to about 20 g/day does not increase the risk of hypertension among women, but beyond this level, the risk increases progressively.