Transfusion of packed red blood cells is common during resuscitation of critically ill patients. However, the association between in-hospital mortality and blood transfusion among patients with severe sepsis during the first 24 hours of hospitalization has not yet been determined. A cohort study was conducted of adult nontrauma patients who visited the emergency department of a tertiary hospital and were diagnosed with severe sepsis. Propensity score (PS) matching was conducted, based on patient demographics, underlying illnesses, laboratory results, and vital signs presented at the emergency department, and multivariate logistic regression was performed to adjust for potential residual confounding between the 2 transfused and nontransfused groups to assess the risk of in-hospital mortality. Of 3448 patients included in this study, 265 underwent blood transfusion during the first 24 hours of hospitalization. Despite comparable severity of sepsis, patients who received transfusions tended to have lower mean arterial pressures (86 vs 98 mmHg) and hemoglobin levels (7.6 vs 11.2 g/dL), and were more likely to have chronic kidney disease (12% vs 6%) and hematologic organ dysfunction (57% vs 35%, all P < 0.001). Transfused patients tended to have higher mortality rates (26% vs 9%, respectively, P < 0.001). After PS matching, 177 pairs of transfused and nontransfused patients were analyzed. After adjusting for residual confounding factors by multivariate logistic regression in the matched patient pairs, no significant differences in in-hospital mortality were observed (odds ratio [OR] = 1.52, 95% confidence interval: 0.92-2.51). In this PS-matched cohort study of adult nontrauma patients with severe sepsis, the in-hospital mortality rate was not significantly different in patients who received blood transfusions during the first 24 hours of hospitalization.