Coagulation studies, i.e. platelet count, prothrombin time (PT) and activated partial thrombin time (aPTT) are commonly employed preoperatively to identify patients at risk. In a retrospective study we evaluated the usefulness of these screening tests to predict postoperative bleeding in 1447 patients with abdominal and thoracic surgery. Forty-six patients (3.2%) experienced postsurgical bleeding. 12.2% of our patients had abnormal coagulation studies. The sensitivity of abnormal coagulation studies with respect to postoperative bleeding was 23.9%. The sensitivity of the parameter "patient at risk", i.e. patients with suspected coagulopathies due to drugs or disease of the liver or kidney, was 56.5%. Thirty-four out of 1008 patients without risk factors had abnormal coagulation tests but an uneventful postoperative course. Preoperative screening of PT and aPTT should be reserved for patients with known or suspected inherited or acquired coagulopathies.