Randomized controlled trials (RCTs) of the relationship between allogeneic transfusion and postoperative bacterial infection at any site have generated discordant results, but have suggested an association of transfusion with wound infection. To examine the specific association of perioperative transfusion with wound infection, we reviewed the records of 964 consecutive patients undergoing elective colorectal cancer resection at our hospital. Diagnoses of wound infection were made retrospectively by the Centers for Disease Control criteria, and transfusion was defined as number of units of white-cell-containing allogeneic blood components received. The probability of wound infection in association with the transfusion was calculated following adjustment for the effects of 12 confounders that had not been previously considered in combination. These factors related to severity of illness, difficulty of operation and risk of wound infection. Wound infection developed in 39 (11.4%) transfused patients, as compared to 24 (3.9%) untransfused subjects (P < 0.0001). In the multivariate analysis, there was a trend suggesting an adverse transfusion effect, which amounted to a 7% increase in the risk of wound infection per unit of red cells or platelets transfused (relative risk [RR] = 1.07; 95% confidence interval for RR = 0.98-1.16; P = 0.1241). This marginally significant effect was related to post-operative (P = 0.1274), rather than perioperative (P = 0.3061), transfusion. We conclude that allogeneic transfusion may perhaps be associated with a small increase in the risk of post-operative wound infection, but this small effect can be established only by RCTs enrolling several thousand patients. The modest magnitude of the effect may have been responsible for the disagreements among the published RCTs.