Autologous transfusion techniques have been proposed to decrease viral contamination related to allogeneic transfusion. Epidemiological studies looking at the actual effect of such policy are scarce. The aim of this study was to assess factors associated to the transfusion of additional allogeneic units among autologous blood recipients in France. We used a subgroup of data of RECEPT, a representative study of transfusion in French hospitals, and described blood recipients and components during the first 24 hours following a transfusion in 175 teaching and non teaching hospitals. We analysed data regarding patients included in the original study, who received autologous blood transfusion (376 patients). Almost one fifth of patients transfused with autologous blood was exposed to additional allogeneic transfusion. Patients undergoing cardiovascular surgery and who were transfused with blood salvage received allogeneic transfusion more frequently than patients undergoing orthopaedic surgery. Further implication of epidemiologists in studies of transfusion or other health technologies could help better formulate policies purported at decreasing risks.