OBJECTIVETo assess the safety and potential efficacy of a mechanical ventilation strategy designed to reduce stretch-induced lung injury in acute respiratory distress syndrome. DESIGNProspective, randomized, controlled clinical trial. SETTINGEight intensive care units in four teaching hospitals. PATIENTSFifty-two patients with acute respiratory distress syndrome. INTERVENTIONSTraditional tidal volume patientstidal volume 10-12 mL/kg ideal body weight, reduced if inspiratory plateau pressure was >55 cm H2 O (7.3 kPa). Small tidal volume patientstidal volume 5-8 mL/kg ideal body weight, to keep plateau pressure < 30 cm H2 O (4.0 kPa). MEASUREMENTS AND MAIN RESULTSMean tidal volumes during the first 5 days in traditional and small tidal volume patients were 10.2 and 7.3 mL/kg, respectively (p < .001), with mean plateau pressure = 30.6 and 24.9 cmH2 O (3.3 kPa), respectively (p < .001). There were no significant differences in requirements for positive end-expiratory pressure or FIO2, fluid intakes/outputs, requirements for vasopressors, sedatives, or neuromuscular blocking agents, percentage of patients that achieved unassisted breathing, ventilator days, or mortality. CONCLUSIONSThe reduced tidal volume strategy used in this study was safe. Failure to observe beneficial effects of small tidal volume ventilation treatment in important clinical outcome variables may have occurred because a) the sample size was too small to discern small treatment effects; b) the differences in tidal volumes and plateau pressures were modest; or c) reduced tidal volume ventilation is not beneficial. (Crit Care Med 1999; 27:1492-1498)