The hypothesis was tested whether retrograde autologous priming (RAP) of the cardiopulmonary bypass system, compared to a standard primed system (NON-RAP group), results in less haemodilution and less transfusion of packed red blood cells. Retrospectively, data was collected from the medical charts of one hundred patients undergoing elective coronary artery bypass grafting using cardiopulmonary bypass. Fifty patients where RAP was used have been compared to fifty patients using NON-RAP. The prime volume in the NON-RAP group was 1,627±108 mL versus 782±96 mL in the RAP group (p<0.001). The lowest haematocrit during perfusion was 22% in the NON-RAP group versus 26% when the RAP technique was used (p<0.001). In the NON-RAP group, 26% of the patients received packed red cells in contrast to 6% in the RAP group (p<0.012). A positive association was found between RAP and less transfusion of packed red blood cells (p<0.012). In conclusion, retrograde autologous priming, reducing the prime volume of the cardiopulmonary bypass system, causes less haemodilution and reduces intraoperative transfusion of packed red blood cells.