The defibrinating agent ancrod has had limited clinical trial, but appears to give no advantages over heparin. Intravenous infusion of dextran, a glucose polymer, has been shown to have an antithrombotic effect in many experimental models of thrombosis. However, the evidence that dextran is a clinically valuable antithrombotic drug is conflicting. A number of controlled randomized studies have shown that dextran can prevent postoperative venous thromboembolism when a large volume of dextran 40 or 70 was infused rapidly during and after surgery. However, blood volume expansion during dextran treatment prohibits its use in patients with reduced cardiac reserve, and infrequent though sometimes severe, allergic reactions have been reported. Evidence that dextran is of value for the treatment of venous or arterial thromboembolism comes from uncontrolled studies and is not convincing. Many compounds have been shown to inhibit platelet function in vitro but only five of these drugs have been extensively evaluated as prophylactic or therapeutic antithrombotic agents in man. These are aspirin, sulphinpyrazone, dipyridamole, hydroxychloroquine and clofibrate. They have been evaluated mainly in patients with cerebral vascular disorders, coronary artery disease, peripheral artery ischaemia, venous thromboembolism, prosthetic heart valves, and in patients with arteriovenous shunts. The evaluation of the clinical effect of the platelet function suppressing drugs is in its early stages, but they appear to differ from each other in the spectrum of their clinical effectiveness, and they may be more effective in arterial than in venous thromboembolic disorders. Their role in the management of cerebral vascular disease and coronary artery disease is still uncertain, and should be clarified by the results of a number of multi-centre, prospective, randomized studies which are currently in progress. Three types of thrombolytic drugs have been evaluated clinically; the plasminogen activators streptokinase and urokinase, proteolytic enzymes such as plasmin, and agents which increase the level of endogenous plasminogen activator (e.g. anabolic steroids). Of these, the plasminogen activators now have a definite place in clinical practice. The plasminogen activators accelerate the lysis of recent venous thrombi and pulmonary emboli, and of arterial thrombi or emboli. Thrombolytic therapy with these agents should be considered particularly in patients with recent major pulmonary embolism, as lysis of recent emboli is rapid and substantial. It should also be considered in patients with recent extensive venous thrombosis, because total lysis of venous thrombi has been reported to result in long-term preservation of valve function, and is likely to prevent postphlebitic syndrome, though this has not been proven. However, plasminogen activator therapy carries a higher risk of bleeding than heparin treatment...