In a randomised double-blind controlled trial 24 patients undergoing elective aortic bifurcation graft surgery received subcutaneous calcium heparin (2, 500 u preoperatively then 5,000 u 12 hourly or 7 days) and 25 control patients received saline injections. All patients received the routine dose of intravenous sodium heparin intraoperatively. The trial was terminated because of excess bleeding complications in patients on subcutaneous heparin (8 vs. 1, p<0.05). Deep venous thrombosis was diagnosed by 125I-fibrinogen scanning in 8 control patients and 3 patients on heparin (p>0.05). In this group of patients the risk of bleeding due to subcutaneous heparin appeared to outweigh the benefit of thrombotic prophylaxis.