A series of 80 patients operated for total hip prosthesis under epidural anesthesia was randomly allocated to treatment with Kabi 2165 (n = 40): 2,500 U anti-Xa preoperatively and evening of operation and 2,500 U anti-Xa morning and evening daily up to the 9th or 10th day postoperatively, or standard heparin (n = 40): 3,750 U preoperatively and then 8 hourly, at a dose adjusted with thrombin time and cephalin + activator time, daily up to the 9th or 10th day. Phlebography was performed routinely on the 9th or 10th day. Venous thrombosis occurred in 7 patients (17.5%) in the Kabi 2165 group, including two high, potentially emboligenic, localizations (5%), and in 4 patients (10%) in the standard heparin group, including 2 potentially emboligenic clots (5%). The difference is not statistically significant (total number: p = 0.5; potentially emboligenic: p = 0.33). Pulmonary embolism did not occur. Overall tolerance, evaluated from hemoglobin and hematocrit values, intra- and post-operative bleeding and operation wound hematoma and at injection site was comparable in the two groups.