In a randomised trial we compared the effects of two different antithrombotic regimens on the incidence of venographically established deep venous thrombosis (DVT) in 83 patients undergoing surgery for fracture of the femoral neck. Group A received dextran 40 peroperatively plus 0.5 g aspirin a day beginning before operation and continuing for 10 days after. Group B received heparin calcium 5000 iu subcutaneously plus dihydroergotamine (DHE) 0.5 mg intramuscularly, given 8-hourly, beginning before operation and continuing for 10 days after. Two patients in Group A and three in Group B developed proximal DVT, while the incidence of all DVT was 33% in Group A and 29% in Group B, a difference which was not significant. Haemorrhagic complications were much more common in the dextran/aspirin group: the volume of drainage fluid, the number of patients transfused and quantity of blood transfused, and the drop in haemoglobin level were all significantly greater in Group A. We conclude that the DHE/heparin regime is preferable to dextran/aspirin because it is safer and no less effective.