The sepsis syndrome represents a systemic response to infection and is defined as hypothermia (temperature 1010F), tachycardia (>90 beat/min), tachypnea (>20 breath/min), clinical evidence of an infection site and with at least one end-organ demonstrating inadequate perfusion or dysfunction expressed as poor or altered cerebral function, hypoxemia (Pao2 <75 torr), elevated plasma lactate, or oliguria (urine output <30 ml/h or 0.5 ml/kg body weight. h without corrective therapy). One hundred ninety-one patients with the sepsis syndrome were evaluated prospectively and comprised the placebo group of a multicenter trial of methylprednisolone in sepsis syndrome and septic shock. Forty-five percent of the patients were found to be bacteremic. Thirty-six percent of the patients were in septic shock (sepsis syndrome plus a systolic BP 40 mm Hg) on study entry. An additional 23% of the patients developed shock after admission with 70% doing so within 24 h of study entry. Shock reversal occurred with a 73% frequency. Twenty-five percent of the patients developed the adult respiratory distress syndrome (ARDS). Mortality for the patients with sepsis syndrome who did not develop shock was 13%. Mortality for the groups of patients with shock on admission and shock postadmission was 27.5% and 43.2%, respectively. Forty-seven percent of the bacteremic patients developed shock after study admission compared to 29.6% of the nonbacteremic patients (p < .05). Aside from development of shock, there were no significant differences between the bacteremic and nonbacteremic patients. The outcomes for the patients with Gram-negative and Gram-positive bacteremias also did not differ significantly. The criteria used to describe the sepsis syndrome in this paper identify a patient population at imminent risk for development of septic shock, ARDS, and death. The identification and definition of this syndrome may allow for earlier detection and treatment of a group of very high-risk patients with sepsis.