The Pediatric Acute Respiratory Failure (PARF) Study Group was organized to study the effects of new therapies on the outcome of lung injury in children. Each center agreed to use similar pressure-limited ventilator strategies including high frequency ventilation (HFV) in all patients. Methods: All intubated children in participating centers with lung injury requiring at least 50% FiO2 and an end expiratory pressure greater than 5cm H2O were enrolled and data were prospectively collected. Patients with significant co-morbidities (chronic lung disease, immunocompromising conditions, cardiac disease, profound, acute central nervous system injury) were excluded from this analysis. Outcome was compared with a validated mortality prediction tool(Crit Care Med 24:323-9, 1996) using data from the Pediatric Critical Care Study Group multi-insitutional study in 1991 (Chest 108: 789-97, 1995) in which patients were selected using the same criteria. Results: In an interim analysis of 131 patients from 7 centers with complete data, 77 patients did not have co-morbidities. Mortality (18.2%) was significantly lower (P=.003) than predicted (32.8%). Improvement appeared in each mortality risk quartile. Severity of disease, as measured by oxygenation index (22 +/- 16) and PRISM score(14 +/- 8), was similar to values from the 1991 predictive data set. The highest peak airway pressure (33 +/- 6 cm H2O) and pCO2 (66 +/- 16 Torr) confirmed the use of pressure limited strategies. The use of high frequency ventilation (34%) was higher than in 1991, but improvement in mortality over predicted was noted in both patients treated with (P=.05) and without (P=.03) HFV. Conclusions: Survival following acute lung injury in patients without co-morbid conditions has improved since 1991 co-incident with the widespread use of pressure limited ventilator strategies.(Supported in part by Ronald MacDonald Charities).