From October 1987 through June 1988 594 patients admitted to the department of visceral surgery, were included in a prospective, randomized double-blind study dealing with postoperative thromboembolic complications. 5000 I.U. heparin were administered twice daily to one group, the other group received a single dose of low molecular heparin (3000 I.U. anti-Xa) daily. In each group more than 290 patients were included. Diagnostic screening was done using LCCT (Liquid Crystal Contact Thermography) and phlebography. 130 patients had to be excluded from the study. 66 patients because of contra-indication, 54 patients because of administrative failing and 10 patients were excluded by the surgeon himself. There was no significant difference between the two groups concerning thromboembolic complications. However analysis of the patients who were excluded from the study shows an increase of thromboembolic complications. Looking at cases of postoperative haemorrhage we found that there was no significant difference between both groups. Interestingly, even compared to the excluded group, there was no difference. In our study postoperative haemorrhagic complications did not depend on the method of prophylaxis but on the surgeon performing the operation. Should a patient have a contra-indication for administration of thromboembolic prophylaxis, he has to take a certain risk of thrombosis. The fact that prophylaxis was not initiated because of neglect or fear of haemorrhagic complications is not only unforgiveable, but dangerous as well.