A case-control study will be carried out with the main objective of evaluating whether the use of noncontraceptive estrogens influences the incidence of first myocardial infarction (MI) in women aged 50-69 years. The effects of recent use (in the month before the MI), past use, the duration of use, and different drug regimens will be evaluated. The study will be large enough to detect reductions in the incidence of about 50%, and increases of the order of 65%, in recent users; reductions of 30% and increases of 40% in past users; and to rule out similar reductions or increases if there is no association. The cases will be identified in 53 hospitals located in greater Boston, a network used recently for a study of MI in men. The controls will be neighbors of the cases identified from town lists. The study will be confined to women 50 to 69 years of age, an age group in which use of noncontraceptive estrogens is common and morbidity and mortality from MI is high. Cases will be identified by weekly telephone calls to the coronary care units. After discharge, standard interviews of the cases and age-matched neighbor controls will be conducted. Information will be collected on use of noncontraceptive estrogens, and on factors that might confound or modify the relation of MI risk to estrogen use: these include correlates of estrogen use (e.g. age at and type of menopause) and MI risk factors (e.g. cigarette smoking). There will be about 900 cases and 900 controls interviewed. Previous studies had not been able to rule out the possibility that apparent decreases in MI risk among estrogen users were due to selection to estrogen use of women who have more healthy life-styles than nonusers. The use of neighbor controls in the proposed study, coupled with information on socioeconomic status, use of medical care, and reasons for starting and stopping estrogen use, will enable more complete control of potential confounding from this source. Previous studies also lacked detailed information on the timing and duration of use, and the drug regimen. The present study will collect and evaluate data relevant to these issues. Insofar as noncontraceptive estrogens are widely used, and use is increasing, and since MI is an important cause of illness and death in older women, an effect of these drugs on the risk of MI would be of major public health importance.