INTRODUCTION: Near-infrared spectrometry assesses hemoglobin saturation of oxygen in tissues (StO2). Although it may provide additional information on local microcirculation function, the usefulness of near-infrared spectrometry in septic patients is debated. This study evaluated whether baseline StO2 value is useful in septic patients admitted to the emergency department with a diagnosis of severe sepsis. MATERIALS AND METHODS: We carried out a prospective multicenter study in three emergency departments in Paris, France. Triage nurses were to record StO2, the results were blinding to the emergency physicians. Patients were eligible when presenting with at least two of the following parameters: temperature higher than 38°C or less than 36°C, respiratory rate greater than 30/min, heart rate greater than 120/min, and systolic arterial blood pressure less than 110 mmHg. Patients with a final diagnosis of sepsis and severe sepsis were analyzed. RESULTS: We analyzed 98 patients (70 with sepsis and 28 with severe sepsis). Thirty-day mortality was 2.9 versus 14.3% (P=0.048) in the sepsis and the severe sepsis group, respectively. No significant difference in the median StO2 was observed in patients with sepsis and severe sepsis [79% (74-85%) vs. 77% (72-83%), respectively; P=0.66]. The area under the curve of the receiver operating characteristic curve for StO2 to predict severe sepsis was 0.53 (0.39-0.66; P=0.5) and the cutoff value was 77%. Median StO2 did not differ in patients admitted to the ICU [80% (60-88%) vs. 79% (74-84%); P=0.78] and in nonsurvivors compared with that of survivors [79% (74-85%) vs. 76% (73-83%); P=0.64]. CONCLUSION: This study fails to show any value of StO2 baseline at triage for early detection of severe sepsis in emergency patients.