We studied the effect of estrogen use on morbidity from cardiovascular disease in 1234 postmenopausal women, aged 50 to 83 years, participating in the Framingham Heart Study's 12th biennial examination (index examination). The medication history recorded at biennial examinations 8 through 12 was used to classify the degree of estrogen exposure before eight years of observation for cardiovascular morbidity and mortality. Despite a favorable cardiovascular risk profile and control for the major known risk factors for heart disease, women reporting postmenopausal estrogen use at one or more examinations had over a 50 per cent elevated risk of cardiovascular morbidity (P less than 0.01) and more than a twofold risk for cerebrovascular disease (P less than 0.01) after the index examination. Increased rates for myocardial infarction (P less than 0.05) were observed particularly among the estrogen users who smoked cigarettes. Conversely, among nonsmokers estrogen use was associated only with an increased incidence of stroke (P less than 0.05). No benefits from estrogen use were observed in the study group; in particular, mortality from all causes and from cardiovascular disease did not differ for estrogen users and nonusers.