In a prospective, randomized, assessor-blind multicentre study two antithrombotic subcutaneous regimens were compared in patients undergoing total hip replacement. Group 1 (154 patients) received 750 anti-Xa units of a new low molecular weight heparinoid (Lomoparan) subcutaneously twice a day and group 2 (155 patients) received 5000 units heparin and 0.5 mg dihydroergotamine (heparin-DHE 5000) twice a day. The incidence of deep vein thrombosis, assessed by routine bilateral venography on day 10 (+/- 1), was 17 and 32 per cent in groups 1 and 2 respectively (risk reduction 47 per cent; P = 0.007). One patient in each group developed a symptomatic pulmonary embolism confirmed by lung scanning. Major bleeding complications occurred in one patient in each group and no significant difference was observed between the two groups with respect to minor bleeding complications. Subcutaneous Lomoparan appears to be as safe as heparin-DHE 5000 at the above doses with regard to bleeding complications, and is more efficacious with respect to venous thrombosis.