OBJECTIVE: Few studies have shown that aged packed red blood cells (RBC) transfusion negatively influenced the outcome of ICU patients, probably related to storage lesions which could be decreased by leukodepletion of RBC. The purpose of this study was to evaluate the impact of aged leukodepleted-RBC pack, on the outcome of ICU patients. DESIGN: Retrospective, observational, cohort study in a Medical Intensive Care Unit. PATIENTS: Consecutive patients admitted during the years 2005 and 2006, and requiring a transfusion. We recorded patient's demographic data, number of RBC unit and age of each RBC, length of ICU, mortality during ICU stay. RESULTS: Five hundred and thirty-four patients were included with global mortality was 26.6%, length of stay in ICU six days (3-14) and SAPS II 48 (35-62). RBC equaling to 5.9 were transfused per patients (22.7%<14 days and 57.3%<21 days). The number of RBC was significantly higher in the dead patients group, but the rate of RBC stored less than 21 days was not different (54% versus 60%; p=0.21). In a multivariate logistic model, independent predictors of ICU death were SAPS II (OR=1.02 per point, p<0.001), number of RBC (OR=1.08 per RBC, p<0.001), length of stay in ICU (p<0.001). Similar results were obtained while introducing the age of RBC as time dependent covariates in a multivariate Cox's model. CONCLUSIONS: RBC transfused in our ICU are old. The ICU outcome is independently associated with the number of leucodepleted RBC transfused, but not with their age.