Totally, 150 patients, subjected to total hip arthroplasty, were randomly allocated into three prophylactic groups with either conventional dextran alone or with additional graded compression stockings or with additional preoperative administration of dextran. The overall frequency of deep venous thrombosis (DVT), as studied by radioactive fibrinogen uptake test and ascending phlebography of the operated on thigh was in the conventional dextran group 46 percent, the additional stockinged group 30 percent, and in the additional preoperative dextran group 52 percent. In the stockinged group, there was a lower frequency of DVT in the nonoperated on leg as well as, on an average, about 350 mL less peroperative bleeding as compared with the other two groups. No adverse reaction occurred from dextran administration. Increased and prolonged postoperative administration of dextran decrease the number of femoral DVTs.