Recent studies have evidenced an association between the time of heparin discontinuance and coronary artery reocclusion. Some investigators have concluded that inadequate heparinization was responsible, and further heparin infusion or an increase in dose would have been indicated. However, several investigators (Rao: Thrombosis Research 24:181-186, 1981; Marciniak and Gockerman: Lancet September 17:581-584, 1977; Fisken et al.: Lancet December 10: 1231, 1977; Conard et al.: Thrombosis Research 22:507-511, 1981; Kakkar et al.: Lancet January 12:103-104; Green: Lancet February 16:374, 375; Harborne and Nicolaides: Thrombosis Research 43:657-662, 1986; Bonen et al.: Thrombosis Research 27:123-124, 1982; Blomback et al.: Acta Physiologica Scandinavica 58:306-318, 1963; Holm et al.: Scandinavian Journal of Haematology 35:564-569, 1985; Andersson et al.: Thrombosis Research 34:333-340, 1984) have demonstrated that antithrombin III levels are reduced in patients on intravenous heparin. Both reduced antithrombin III levels and reduced rate of antithrombin inhibition of thrombin at the time of heparin discontinuance may increase the risk of coronary rethrombosis. If this theory is correct, increasing heparin infusion may exacerbate this risk. We propose an investigation that will provide evidence for or against the decreased antithrombin III theory, and in doing so, test an experimental therapy designed to prevent coronary reocclusion upon heparin discontinuance. In a randomized, placebo-controlled double-blinded study, we will determine whether simultaneous administration of warfarin with heparin initiation provides more time to increase antithrombin III levels and prevent coronary reocclusion upon heparin discontinuance, compared to heparin without warfarin therapy.