Sixty-seven hip-arthroplasty and fifty-two hip-fracture patients participated in a placebo-controlled randomized double-blind study on the effects of low-dose heparin prophylaxis in the prevention of venous thromboembolism. In this study, a positive thromboembolic event meant a positive test by: (1) daily 125I-fibrinogen scanning, (2) contrast venography on the tenth postoperative day, or (3) radionuclide perfusion lung scan in confirmation of suspected clinical pulmonary emboli. Nineteen (59.4 per cent) of thirty-two placebo-treated arthroplasty patients showed evidence of a thromboembolic event in contrast with eight (22.9 per cent) of thirty-five heparin-treated patients (p less than 0.003). Heparin-treated arthroplasty patients required mean blood transfusions of 4.7 units, contrasted with a mean 3.2-unit transfusion requirement for placebo-treated patients (p less than 0.05). The incidence of observed bleeding complications was higher among the heparin-treated patients. Of the twenty-three placebo-treated patients with fracturs, 39.1 per cent had a thromboembolic event, while 41.4 per cent of the twenty-nine who received heparin showed evidence of thromboembolism, demonstrating that low-dose heparin afforded no protection, nor did it affect the incidence of bleeding complications or transfusion requirements in fracture patients.