Eight thousand and one patients were admitted to a prospective, randomized multicentre trial comparing the prophylactic efficacy of heparin-dihydroergotamine (DHE) and Dextran 70 against fatal pulmonary embolism after emergency or elective orthopaedic surgery. Of the 7413 patients correctly admitted to the trial, 3698 were allocated to receive heparin-DHE and 3715 to receive Dextran 70. Twenty-eight patients in the heparin-DHE and 27 in the dextran group died within 40 days of operation. Necropsy was performed in 19 cases in each group. In 6 patients in the heparin-DHE and in 9 in the dextran group pulmonary embolism was the sole or a contributory cause of death (n.s.). In the dextran group there were more diffuse intraoperative bleeding complications (P less than 0.01), more massive postoperative haemorrhages (P less than 0.01), more small wound haematomas (P less than 0.05) and more allergic reactions (P less than 0.05). This is the first large multicentre trial dealing with elective and emergency procedures in orthopaedic patients.