One-hundred-fourteen consecutive cases of venography after primary uncemented total hip replacement were performed in a randomized trial to identify the natural incidence of deep vein thrombosis, the effectiveness of prophylactic regimens such as 1.2 grams of aspirin daily and low-molecular-weight dextran for 3 days, and other relative factors for the development of venous thrombosis. In addition, intraoperative venography was conducted in 10 patients to study the speed of the flow of contrast media in the femoral vein and the development of deep vein thrombosis and the extent of the twisting of the femoral vein during hip joint manipulation. The incidence of venous thrombosis in the control, aspirin, and dextran prophylaxis groups were 20%, 11.5%, and 5.2% respectively. The incidence in the aspirin group was reduced, but this was statistically insignificant. The dextran group showed a marked decrease in incidence, and the difference with the control group was statistically significant. With regard to the development site of venous thrombosis, it was prevalent in iliofemoral, lower femoral, and major calf vein in the control group, while the popliteal and major calf vein were the major site of thrombosis development in the aspirin and dextran groups. The risk factors affecting the incidence of the venous thrombosis are confirmed to be obesity and long-term administration of steroids. Hematologic analysis was meaningless in investigating the development of venous thrombosis. The reliable clinical sign and symptom suggestive of the development of venous thrombosis was the severe swelling on the entire portion of affected lower extremity. In the intraoperative venogram, no correlation was found between the venous blood flow speed and the development of venous thrombi. A remarkable change in the blood flow of the femoral vein was noticed when the hip joint was flexed an average of 40.4 degrees, adducted at 11.5 degrees, and internally rotated at 81.5 degrees. Especially, when the joint was internally rotated, severe kinking of the vein could be observed. Thus it seems desirable to reduce the duration of internal rotation of the hip joint as much as possible to prevent venous thrombosis.