Of the various prophylactic agents evaluated, four have been found to be effective. These are oral anticoagulants, low-dose heparin, mechanical devices which increase venous blood-flow in the leg, and Dextran. Oral anticoagulants have been shown to be effective in patients having abdominal, thoracic, or hip surgery, when treatment was started either before surgery or in the immediate postoperative period. They have also been shown to be effective in medical patients. The evidence derives from studies which showed that treatment can reduce total mortality, prevents venous thromboembolism detected clinically or at autopsy, and prevents thrombosis diagnosed with venography. On the other hand, the incidence of thrombosis diagnosed by 125I-fibrinogen scanning was not reduced when oral anticoagulants were started just before or just after surgery. This suggests that oral anticoagulant treatment starting in the immediate postoperative period may not prevent formation of the initial thrombotic nidus, but is clinically effective because it prevents extension of the nidus to form a significant thrombus. Bleeding has been a significant complication in almost all studies of surgical patients, and this is the major factor which has prevented widespread use of oral anticoagulant prophylaxis. In addition, the need for careful laboratory monitoring makes this approach inconvenient and adds to its expense. Low-dose heparin has been shown to be effective in general surgical and medical patients, but results have been inconclusive in patients having elective hip surgery, and this approach is probably ineffective in patients with hip fracture. In general surgical patients, low-dose heparin prophylaxis has been shown to prevent pulmonary embolism diagnosed at autopsy examination or with lung scanning, and calf and thigh vein thrombosis diagnosed with 125I-fibrinogen leg-scanning. A slight, but statistically significant, increase in the frequency of wound hematoma and a greater postoperative hematocrit fall have been reported when heparin was given three times daily, but not with the twice daily heparin injection regimen. In these studies, low-dose heparin was given without laboratory control of its anticoagulant effect, so that this prophylactic approach is simple, but the need for subcutaneous injections is a disadvantage of this approach. Results with methods which increase venous blood-flow in the leg have varied, depending on the technic used. Active measures, such as intermittent pneumatic calf compression or peroperative electrical calf muscle stimulation, have been shown to prevent thrombosis detected with 125I-fibrinogen leg-scanning. However, while the evidence suggests that both methods are effective in relatively low risk patients, they may have limited value in the high risk patient who is confined to bed for a long time. These methods are free of side effects and relatively inexpensive, but intermittent calf compression, in particular, is slightly cumbersome...