Objective: To evaluate the safety and effectiveness of high-frequency oscillatory ventilation using a protocol designed to recruit and maintain optimal lung volume in patients with severe adult respiratory distress syndrome (ARDS). Setting: Surgical and medical intensive care units in a tertiary care, military teaching hospital. Design: A prospective, clinical study. Patients: Seventeen patients, 17 yrs to 83 yrs of age, with severe ARDS (Lung Injury Score of 3.81 +/- 0.23) failing inverse ratio mechanical conventional ventilation (Pao(2)/Fio(2) ratio of 68.6 +/- 21.6, peak inspiratory pressure of 54.3 +/- 12.7 cm H2O, positive end- expiratory pressure of 18.2 +/- 6.9 cm H2O). Interventions: High-frequency oscillatory ventilation was instituted after varying periods of conventional ventilation (5.12 +/- 4.3 days). We employed a lung volume recruitment strategy that consisted of incremental increases in mean airway pressure to achieve a Pao(2) of >60 torr (>8.0 kPa), with an Fio(2) of <0.6. Measurements and Main Results: High frequency oscillator ventilator settings (Fio(2), mean airway pressure, pressure amplitude of oscillation [Delta P] frequency) and hemodynamic parameters (cardiac output, oxygen delivery [Do(2)]), mean systemic and pulmonary arterial pressures, and the oxygenation index (oxygenation index = [Fio(2) x mean airway pressure x 100]/Pao(2)) were monitored during the transition to high-frequency oscillatory ventilation and throughout the course of the high-frequency protocol. Thirteen patients demonstrated improved gas exchange and an overall improvement in Pao(2)/Fio(2) ratio (p < .02). Reductions in the oxygenation index (p < .01) and Fio(2) (p < .02) at 12, 24, and 48 hrs after starting high frequency oscillatory ventilation were observed. No significant compromise in cardiac output or Do(2) was observed, despite a significant increase in mean airway pressure (31.2 +/- 10.3 to 34.0 +/- 6.7 cm H2O, p < .05) on high-frequency oscillatory ventilation. The overall survival rate at 30 days was 47%. A greater number of pretreatment days on conventional ventilation (p < .009) and an entry oxygenation index of >47 (sensitivity 100%, specificity 100%) were associated with mortality. Conclusions: High frequency oscillatory ventilation is both safe and effective in adult patients with severe ARDs failing conventional ventilation. A lung volume recruitment strategy during high-frequency oscillatory ventilation produced improved gas exchange without a compromise in Do(2). These results are encouraging and support the need for a prospective, randomized trial of algorithm-controlled conventional ventilation vs. high frequency oscillatory ventilation for adults with severe ARDS.