INTRODUCTION: Derangements of haemostasis and associated blood loss still represent a problem in cardiac surgery. The synthetic vasopressin analogue desmopressin acetate (DDAVP, Minirinq has been found to decrease blood loss and homologous blood requirement in complicated cardiac operations (1). But these results could not be confirmed in more recent studies, neither in adults nor in children requiring cardiac surgery (2,3). DDAVP is thought to ameliorate platelet function by increasing factor VIII: von Willebrand factor (F VIII) activity (1,4). Since preoperative antiplatelet therapy increases postoperative blood loss and transfusion requirements (51, we investigated the bloodsaving effect of DDAVP in this group of cardiac surgical patients. PATIENTS AND METHODS: With Ethical Committee's approval, and after informed consent was obtained, 40 consecutive male patients requiring coronary artery bypass grafting (CABG) were assigned to the double-blind, placebo-controlled study. All patients had taken aspirin or aspirin-like drugs until at least 5 days before the operation. Operations were performed using standard anaesthetic, perfusion and anticoagulation techniques. After termination of CPB and administration of protamine sulfate (4.5 mg/ kg), patients randomly received an infusion of either 0.3 ug/ kg DDAVP (group D) or an equivalent amount of saline (group C) during 15 minutes. Intra-and postop. transfusion requirements, blood loss and urine output were registered. The red cell mass (RCM; product of haematocrit and blood volume) was calculated preop., at discharge from the ICU and 7 days postop. The bleeding time (8T) as well as the activated clotting time (ACT) was measured preop., after DDAVP administration and at chest closure. Additionally, in 20 patients (10 of each group) the following parameters were investigated (preop., after protamine application, at chest closure. 4 hours oostoo.): TPZ. PTT. F VIII, tissue-plasminogen activator (tPA), tPA inhibitor, fibrine monomeres and fibrine split products (D-dimers). Statistical analysis was carried out using ANOVA; regarding p do.05 being significant. RESULTS: There was no difference between groups concerning demographic data, duration of operation or CPB or number of grafts performed. The total homologous blood requirement was significantly lower in group D compared to group C (972+197 vs. 1,868+227 ml). Although at all poinFs of measure&t (intra-and postop.) transfusion requirement was less in group D (Fig. 11, haematocrit values in group D always exceeded those of group C. Four hours postop.,