A prospective, randomized, assessor-blind trial has been undertaken to compare the thromboprophylactic effect and safety of the heparinoid Org 10172 (a mixture of low molecular-weight sulfated glycosaminoglycuronides) and dextran 70 in patients operated on for hip fracture. Prestudy biostatistical calculations led to the need for 260 patients. Three hundred eight patients were randomized and 19 were excluded after randomization, the majority because of postponed surgery. Analyses were made on the 289 patients on an intention-to-treat basis, as well as on the 247 patients given correct prophylaxis. Diagnosis of deep vein thrombosis was based on bilateral ascending phlebography on postoperative days 10 through 12. The frequency of deep vein thrombosis on an intention-to-treat basis was 10% in the Org 10172 group and 30% in the dextran 70 group and, on the basis of correct prophylaxis, 12% and 31%, respectively, both differences being significant (p less than 0.001). Two-month mortality rates were equal in the groups. Three fatal pulmonary emboli were seen in the dextran group. Significantly more patients in the dextran group received postoperative transfusions; no other differences in various hemorrhagic parameters were seen. Thus it can be concluded that Org 10172 has a significantly better thromboprophylactic effect than does dextran in patients with hip fractures without significant side effects.