Acute coronary syndromes are caused by coronary artery thrombosis. Given the central role of platelets and coagulation in the pathogenesis of acute coronary syndromes, aspirin and heparin are the cornerstones of current treatment strategies. Despite their widespread use, however, both agents have limitations in this setting. Thus, aspirin fails to block platelet activation by agonists, such as thrombin, which act through thromboxane A2-independent pathways. Arterial thrombosis is relatively resistant to heparin because (a) heparin is neutralised by platelet factor 4 released from platelets activated at sites of plaque distribution and (b) the heparin-antithrombin complex is unable to inactivate thrombin bound to the fibrin clef. New antithrombotic agents have been designed to overcome the limitations of aspirin and heparin. Antagonists of the platelet glycoprotein IIb/IIIa receptor (GP IIb/IIIa) and the direct thrombin inhibitors, hirudin and hirulog, are in the most advanced stages of clinical testing. By blocking the final common pathway of platelet aggregation, GP IIb/IIIa antagonists prevent platelet aggregation by all agonists, including thrombin. Hirudin and hirulog have potential advantages over heparin because they are not neutralised by platelet release products and, unlike heparin, can inactivate thrombin bound to fibrin. Chimeric 7E3 (c7E3), a humanised monoclonal antibody against GP IIb/IIIa, has been shown to reduce recurrent ischaemic events in patients undergoing coronary angioplasty both at 30 days and at six months. Thus, c7E3 is the first agent to produce a clinically meaningful reduction in long-term ischaemic outcomes in a large-scale randomised clinical trial. To date the results with synthetic GP IIb/IIIa antagonists have been less promising, with early benefits being lost over time. This may reflect their shorter half-life compared to c7E3. Hirudi and hirulog appear to be superior to heparin in preventing recurrent ischaemic events early after coronary angioplasty, but this effect is transient indicating that there is ongoing thrombin generation at the site of vascular injury. Taken together these findings suggest that long-term anticoagulant therapy is needed to block ongoing activation of coagulation after early treatment with direct thrombin inhibitors or synthetic GP IIb/IIIa antagonists. This may not be necessary with c7E3 because its biological half-life is so prolonged. Further studies are needed to test these concepts.