The efficacy of two antithrombotic regimens, combined dextran and aspirin and combined dextran and warfarin, was analyzed by comparing the incidence of thromboembolism following total hip replacement in two groups of similar patients. Of the 427 who received dextran and aspirin, 7 per cent had thromboembolic complications, including one case of fatal pulmonary embolus and one case of recurrent emboli that required vena caval ligation, and 15 per cent had wound-healing complications. Of the 197 patients who received dextran and warfarin, 5 per cent had thromboembolism and 24 per cent had wound healing complications. Although both prophylactic regimens seemed effective, dextran and aspirin appeared less effective in reducing thromboembolic complications than dextran and warfarin, but there were fewer wound complications in that group. One-fourth of the patients on dextran-warfarin were not adequately anticoagulated despite close supervision. In forty-five patients with a history of thromboembolism who were excluded from the study and analyzed separately, warfarin alone and the two described regimens were equally ineffective in preventing thromboembolism, and the incidence of thromboembolic complications was high. Dextran-aspirin and dextran-warfarin appear to be satisfactory and relatively simple methods of thromboembolic prophylaxis.