Soon after oral contraceptives (OCs) were first marketed in the 1960s, they were linked to an increased risk of thromboembolic disease. Over the years, the estrogen and progestogen doses have been decreased in the hope of reducing the thrombogenic effect. Studies of OCs used in the 1970s (monophasic OCs containing greater than or equal to 50 microg of estrogen) indicated an increased risk of myocardial infarction (Ml) in current users relative to nonusers, with the effect concentrated in women who smoked heavily. The effects of newer formulations that are currently in use, most with <50 microg of estrogen, have not been established. We are just completing a case-control study of MI designed to assess their effect on MI risk. Gases of first - MI have been identified in a network of some 100 hospitals in Massachusetts, Rhode Island, and Pennsylvania. Based on analyses of 314 cases and 1914 controls studied to date, the RR for current OC users relative to non users is not significantly increased among nonsmokers or light smokers; however, the RR among smokers of greater then or equal to 25 cigarettes per day is 4.0 (95% Cl 0.9-17), and the RR for the joint effect of current OC use and heavy smoking, relative to nonuse and nonsmoking, is 35 (95% Cl 15-78). These results are closely similar to findings in a study of Ml that we conducted in 1976-1979. Thus, the present results do not support the hope that the lower dose OCs are safer in regard to Ml than the older higher dose preparations, nor do other available epidemiologic data. The present results are based mainly on use of monophasic OCs. Recently, triphasic formulations have become popular and they now account for about half of current OC use. There were not enough users of triphasic OCs in the present study for informative analysis, nor are there any data from other studies on this issue. We propose to continue data collection to establish the effect of triphasic OCs on MI risk, with particular attention to the joint effect with heavy smoking. The new data collection will also permit more definitive analysis of the effect of monophasic OCs. A likely mechanism for the effect of OCs on Ml risk is an increased tendency to thrombosis. Persons with blood group A also have an increased risk of clotting. The continuation of data collection will allow us to collect, at no additional expense, data on blood group in the new set of cases and controls. We will assess the independent effect of blood group A on MI risk, and its interaction with cigarette smoking and with OC use. Given the high prevalence of blood group A, interactions of these factors to affect MI risk would be of public health importance.