A surfactant-depletion lung-injury model was produced in 37 New Zealand white rabbits by saline lavage. During the next 2 to 3 h, rabbits were ventilated with conventional mechanical ventilation (CMV, group 1), high-frequency jet ventilation (HFJV, group 2), or CMV for 1 h followed by HFJV for 2 h (CMV/HFJV, group 3). Survival until planned termination of the protocol was 56%, 77%, and 63% in groups 1, 2 and 3, respectively. Causes of early demise were usually pneumothorax or metabolic acidosis. There were no statistically significant differences among the groups with respect to survival, incidence of pneumothorax or metabolic acidosis. Arterial oxygenation was more efficient with HFJV (group 2) (P[A-a]O2 = 372 +/- 51 torr [mean +/- SE] at 2 h) than with CMV (group 1) (P[A-a]O2 = 512 +/- 18 torr at 2 h, p less than .01). Furthermore, oxygen gas exchange in 3 of 5 group 3 rabbits improved after institution of HFJV. In contrast to previous findings with high-frequency oscillation (HFO), there were no qualitative histologic differences between lungs ventilated with HFJV vs. CMV. Thus, although HFJV produced more efficient gas exchange in this model, it did not improve pulmonary pathology. HFO may be preferable to HFJV in infant respiratory distress syndrome.