There is a need for insight into factors that contribute to late mortality of sepsis patients. Immunomodulatory effects have been ascribed to blood transfusion. This retrospective cohort study investigates the association between the development of nosocomial bacterial infection and transfusion of leukodepleted red blood cells (RBCs) or platelets (PLTs) in survivors of the initial phase of sepsis. Patients diagnosed with sepsis after admission to the intensive care unit of a tertiary referral hospital were included. Of 134 patients with sepsis, 67 received a blood transfusion (50%). A secondary infection developed in 19 patients (14%). A multiple logistic regression model revealed that the use of immunosuppressive medication with an odds ratio (OR) of 1.17 (95% confidence interval [CI], 1.04-1.31), but not Acute Physiology and Chronic Health Evaluation II score, malignancy, HIV infection, alcohol abuse, or diabetes mellitus, was a risk factor for nosocomial infection. In an adjusted model, the amount of transfused RBCs was associated with secondary infection with an OR of 1.18 (95% CI, 1.01-1.37). Storage time of RBCs was a relevant confounder of the effect of the amount of RBCs on infection, with an adjusted OR of 1.25 (95% CI, 1.04-1.51), P = 0.02. Also, the amount of transfused PLTs was associated with secondary infection, with an OR of 1.36 (95% CI, 1.05-1.78). In conclusion, transfusion of RBCs and PLTs is associated with the onset of secondary bacterial infection in sepsis patients. Storage time of RBCs influences this increased risk. These findings suggest that immunomodulatory effects of blood transfusion contribute to adverse outcome in the convalescent phase of sepsis.