Background Postoperative mortality is greater in women than men after coronary artery bypass graft surgery. Because allogeneic blood transfusions are more common in women and have been associated with immunomodulation, the impact of transfusion on sex differences in infection and mortality was examined. Methods A cohort study was conducted using Michigan Medicare beneficiaries who had undergone coronary artery bypass graft surgery. Information was used regarding allogeneic blood transfusion, infection, and mortality within the 100-day period after surgery. Results Blood transfusions were more common in women than in men (88.2%, 95% CI 87.1%-89.2% vs 66.7%, 95% CI 65.5%-67.9%). Patients who received transfused blood were more likely to have an infection than patients who did not (14.6%, 95% CI 13.8%-15.5% vs 4.9%, 95% CI 4.1%-5.9%). There was a dose-response relationship between the number of units of whole blood or packed red cells received and the prevalence of infection (P = .035). The unadjusted risk of mortality attributable to female sex was 13.9% (95% CI 8.1%-19.6%), but was no longer statistically significant when adjusted for blood transfusion (population attributable risk 0.6%, 95% CI −6.0% to 6.6%). Patients who received a transfusion were 5.6 times as likely to die within 100 days after surgery as those who did not receive a transfusion (95% CI 3.7-8.6). Conclusion The increased risk of mortality in women after bypass surgery may be explained by transfusion-related immunosuppression.