In a randomized clinical trial the effect of subcutaneous heparin alone or in combination with dihydroergotamine or sulphinpyrazone in preventing postoperative deep vein thrombosis (DVT) was studied. Sodium heparin (5000 IU) was administered subcutaneously twice daily; dihydroergotamine (1/2 mg) was also administered subcutaneously twice daily, and sulphinpyrazone (400 mg) was administered orally or intravenously twice daily. Administration occurred for at least 7 days. The diagnosis DVT was made with the radiofibrinogen uptake test. 358 patients undergoing major elective abdominal surgery were allocated to three treatment groups: heparin alone (Hep), heparin + dihydroergotamine (DHE-Hep) and heparin + sulphinpyrazone (Sulph-Hep). The frequency of DVT was 14/114 in Hep, 10/115 in DHE-Hep and 20/114 in Sulph-Hep. These differences were not significant. After application of the "logistic regression" procedure of Cox (1) it turned out that the major risk factors for developing DVT were age, sex, weight, type of operation and presence of diabetes mellitus. Also a significant treatment influence was observed (p = 0.001). This treatment effect was most probably due to improvement in the DHE-Hep group. The results in the Sulph-Hep group were not significantly different from those in the Hep group. A risk index was formulated on the basis of the above mentioned risk factors by which the chance of occurrence of DVT during heparin prophylaxis in an individual patient could be predicted. Patients that should receive additional prophylactic treatment can be defined by using this risk index.