In a randomized, prospective trial, a low-molecular-weight heparinoid (Org 10172 [Lomoparan]) was compared with warfarin for efficacy and safety in preventing deep-vein thrombosis in 263 patients who had an operatively treated fracture of the hip. One group of patients received Org 10172 in a dose of 750 units subcutaneously every twelve hours until the ninth postoperative day; on the seventh postoperative day, warfarin was added to the regimen. The other group received only warfarin. Both drugs were begun preoperatively, immediately after the admission evaluation. In the patients who received warfarin, the desired prothrombin time was one and one-half times the control level. Deep-vein thrombosis was detected by 125I-fibrinogen scanning and impedance plethysmography and was confirmed by phlebography and compression ultrasonography. Deep-vein thrombosis was found in nine (7 per cent) of the 132 patients who received Org 10172 and in twenty-eight (21 per cent) of the 131 patients who received warfarin (p less than 0.001). Adverse reactions were not significantly different in the two groups. Major bleeding complications occurred in eight patients in the Org-10172 group, only four of whom were receiving the drug at the time of bleeding, and in five patients who were receiving warfarin (not significant). There was no difference in intraoperative loss of blood or in requirements for transfusion. We concluded that the low-molecular-weight heparinoid Org 10172 is a safe, convenient, effective antithrombotic agent for the prevention of venous thrombosis after an operation for fracture of the hip.