We tested the hypothesis that high frequency oscillatory ventilation (HFOV) would result in decreased pulmonary barotrauma in infants with hyaline membrane disease by comparing HFOV at 10 Hz to conventional positive pressure ventilation with continual distending airway pressure (PPV/PEEP) in premature baboons with hyaline membrane disease. Nineteen baboon fetuses were randomized to one of two treatment groups, delivered at 140 +/- 2 days, and, after stabilization and instrumentation of PPV/PEEP, placed in their respective ventilator group. Animals on conventional ventilation were managed by adjustment of tidal volume and frequency (to 1 Hz) to keep PaCO2 below 55 and by adjustment of the mean airway pressure. One of the "HFOV" group died of cardiovascular complications before going on HFOV and was eliminated from data analysis. The remaining HFOV baboons survived the 11-day experimental period without evidence of airleak. Six of the 11 prematures treated with PPV/PEEP developed pulmonary interstitial emphysema and/or pneumothorax and five of the animals died within 48 h. The intergroup differences in airleak were significant (p less than 0.05). Mean airway pressure (measured at the proximal airway) was higher initially with HFOV but then was lowered more rapidly than in the PPV/PEEP animals. The arterial to alveolar oxygen ratio rose and the FIO2 could be lowered more rapidly with HFOV than with conventional ventilation. These differences reached significance by 20 h. After 60 h there were no significant differences between HFOV and the PPV/PEEP survivors. HFOV resulted in more uniform saccular expansion, higher arterial to alveolar oxygen ratio, less oxygen exposure, and decreased acute barotrauma when compared to PPV/PEEP.(ABSTRACT TRUNCATED AT 250 WORDS)