OBJECTIVE Diabetes patients have an increased risk of sepsis. Several inflammatory and coagulant pathways that are activated during sepsis are also up-regulated in diabetes patients. We tested our a priori hypothesis that the presence of diabetes adversely affects the outcome of sepsis. DESIGN Retrospective analysis of a previously published study. SETTING Intensive care units of 164 centers in 11 countries. PATIENTS Eight hundred thirty severe sepsis patients who were admitted to the intensive care unit and who received standard critical care treatment. INTERVENTIONS Patients were stratified into diabetic and nondiabetic patient groups. Mortality was assessed after 28 and 90 days, causative microorganisms were evaluated, and markers of coagulation, fibrinolysis, and inflammation were measured at several time points. MEASUREMENTS AND MAIN RESULTS Diabetes was present in 22.7% of all sepsis patients. Throughout the study, plasma glucose levels were higher in diabetic patients. Mortality was equal in diabetic and nondiabetic patients (31.4% vs. 30.5% after 28 days). Markers of coagulation, fibrinolysis, and inflammation were generally equal in diabetic and nondiabetic patients, although on admission diabetic patients had slightly higher levels of anticoagulation markers. Interestingly, nondiabetic patients with admission hyperglycemia (>11.1 mmol/L; 200 mg/dL) had a higher mortality rate compared to those without admission hyperglycemia (43.0% vs. 27.2%). CONCLUSIONS Although diabetes is a risk factor for sepsis, once established, the outcome of severe sepsis does not appear to be significantly influenced by the presence of diabetes. In nondiabetic patients, however, admission hyperglycemia is associated with an increased mortality.