Trauma patients are at risk for thromboembolic complications, but effective methods of prophylaxis have not been established for this heterogenous population. In this prospective trial, 400 trauma patients were assigned to one of three groups, depending upon their injuries, and randomized within each group to a treatment mode: Group I: sequential gradient pneumatic leg compression (SCD), low-dose subcutaneous heparin (H), or control (C); Group II: H or C; Group III: SCD or C. Venous duplex ultrasound examinations were performed on admission and weekly thereafter. Of the 251 patients who completed the study, 15 (6%) developed lower extremity venous thrombosis and two additional patients developed pulmonary embolism (one fatal). Significant risk factors associated with the development of thromboembolism included immobilization > 3 days, age 30 years or older, and the presence of pelvic or lower extremity fractures. In patients with neurotrauma who cannot receive heparin (Group III), the SCD was more effective than control in preventing DVT (p = 0.057). Neither H nor SCD appeared to offer protection for the other groups of trauma patients, but surveillance with ultrasound examinations allowed for prompt recognition and treatment of occult deep vein thrombosis.