Objectives:Therationaleof usingautotransfusionofmediastinalshedbloodaftercardiacsurgeryis to preservehaemoglobin levelsandreduce the need for allogenic blood transfusions. However, the method is controversial and its clinical value has been questioned. We hypothesised that re-transfusion of mediastinal shed blood instead impairs haemostasis after routine coronary artery bypass grafting and thus increases postoperative bleeding. Methods: Seventy-seven consecutive elective coronary artery bypass surgery patients (mean age 67 9 years, 77% men) were included in a prospective, randomised controlled study. The patients were randomised to postoperative re-transfusion of mediastinal shed blood (n = 39) or to a group where mediastinal shed blood was discarded (n = 38). Primary end point was bleeding during the first 12 postoperative hours. Secondary end points were postoperative transfusion requirements, haemoglobin levels, thrombo-elastometric variables and plasma concentrations of interleukin-6, thrombin—anti-thrombin complex and D-dimer. Results: Mean re-transfused volume in the autotransfusion group was 282 210 ml. There was no difference in postoperative bleeding (median 394 ml (interquartile range 270—480) vs385(255—430)ml,p = 0.69),proportionofpatientsreceivingtransfusionsofbloodproducts(11/39vs11/38,p = 0.95),haemoglobinlevels24 h after surgery (116 13 vs 116 14 g l 1 , p = 0.87), thrombo-elastometric variables, interleukin-6 (219 144 vs 201 144 pg ml 1 , p = 0.59), thrombin—anti-thrombin complex (11.0 9.1 vs 14.8 15, p = 0.19) or D-dimer (0.56 0.49 vs 0.54 0.44, p = 0.79) between the autotransfusion group and the no-autotransfusion group. Conclusions: Autotransfusion of small-to-moderate amounts of mediastinal shed blood does not influence haemostasis after elective coronary artery bypass grafting. # 2010 European Association for Cardio-Thoracic Surgery. Published by Elsevier B.V. All rights reserved.