Résumé : 99 patients who had undergone a major surgical operation were distributed at random into 2 groups: a control group of 55 patients and another group of 44 patients receiving 5000 units of subcutaneous heparin 2 hr before the operation, then 5000 units 3 times daily for 6 days. In the control group, the number of postoperative deep venous thromboses clinically apparent was 7, whereas no case of deep vein thrombosis was diagnosed clinically in the treated group. Deep vein thrombosis was also sought by the technic of labelled fibrinogen in 27 patients of the control group and 23 patients of the treated group. This test was positive in 17 control operated patients and in 6 treated operated patients. The difference is statistically significant. The number of transfusions and the average quantity of blood transfused was identical in both groups. The administration of low doses of subcutaneous heparin prevents effectively and without risk of bleeding, deep vein thrombosis.