BACKGROUND The efficacy of acute normovolaemic haemodilution (ANH) remains uncertain because of a lack of well-designed prospective randomized controlled trials. The aim of this study was to assess the effects of ANH on allogeneic transfusion, postoperative complications, and duration of stay. METHODS Consecutive patients undergoing major gastrointestinal surgery were randomized to a planned 3-unit ANH, or no ANH. Both groups underwent identical management including adherence to a transfusion protocol after surgery. Outcome measures included the number of patients receiving allogeneic blood, complications, and duration of stay. RESULTS 380 patients were screened of which 160 were included in the study, median age was 62 yr (range 23-90), 'ANH' n=78, 'no ANH' n=82. There was no significant difference between groups in the number of patients receiving allogeneic blood 22/78 (28%) vs 25/82 (30%), the total number of allogeneic units transfused (90 vs 93), complication rate, or duration of stay. Haemodilution significantly increased anaesthetic time, median 55 (range 15-90) vs 40 min (range 17-80) (P<0.001). Significantly fewer patients in the ANH group experienced oliguria in the immediate postoperative period 37/78 (47%) vs 55/82 (67%) (P=0.012). The most significant factors affecting transfusion were blood loss, starting haemoglobin, and age. When compared with ASA-matched historical controls, the introduction of a transfusion protocol reduced the transfusion rate in colorectal patients from 136/333 (41%) to 37/138 (27%), P=0.004. CONCLUSIONS In this large pragmatic study, ANH did not affect allogeneic transfusion rate in major gastrointestinal surgery. Preoperative haemoglobin, blood loss, and transfusion protocol are the key factors influencing allogeneic transfusion.