The incidence of thromboembolism was determined by a survey of 7986 orthopedic patients who had 5966 operations at a university hospital during a three-year period. Patients were analyzed to determine thromboembolic risks with respect to the orthopedic diagnosis to assist in providing the patient with informed consent and to identify patients who need prophylactic anticoagulation therapy. The patient population was divided into three risk classes: Class 1, low-risk patients, including children, all upper-extremity patients, and adult non-surgical ambulatory rehabilitation patients; Class 2, medium-risk patients, including all adult lower-extremity patients and patients with spine problems and no history of thromboembolic disease; Class 3, high-risk patients, including all patients with a previous history of thromboembolic disease or venous stasis disease. Class 1 patients require no prophylactic anticoagulation, Class 2 patients should be treated with dextran or aspirin, and Class 3 patients should be treated with warfarin. The calculated mortality for Class 2 patients, excluding hip problems, is 0.26%; mortality for acute spinal cord injuries is 2.9% and for hip surgery 2.2%.