OBJECTIVE: To determine whether elevations in pulmonary vascular pressure induced by mechanical ventilation are more injurious than elevations of pulmonary vascular pressure of similar magnitude occurring in the absence of mechanical ventilation. DESIGN: Prospective comparative laboratory investigation. SETTING: University research laboratory. SUBJECTS: Male New Zealand white rabbits. INTERVENTIONS: Three groups of isolated, perfused rabbit lungs were exposed to cyclic elevation of pulmonary artery pressures arising from either intermittent positive pressure mechanical ventilation or from pulsatile perfusion of lungs held motionless by continuous positive airway pressure. Peak, mean, and nadir pulmonary artery pressures and mean airway pressure were matched between groups (35, 27.4 +/- 0.74, and 20.8 +/- 1.5 mm Hg, and 17.7 +/- 0.22 cm H2O, respectively). MEASUREMENTS AND MAIN RESULTS: Lungs exposed to elevated pulmonary artery pressures attributable to intermittent positive pressure mechanical ventilation formed more edema (6.8 +/- 1.3 vs. 1.1 +/- 0.9 g/g of lung), displayed more perivascular (61 +/- 26 vs. 15 +/- 13 vessels) and alveolar hemorrhage (76 +/- 11% vs. 26 +/- 18% of alveoli), and underwent larger fractional declines in static compliance (88 +/- 4.4% vs. 48 +/- 25.1% decline) than lungs exposed to similar peak and mean pulmonary artery pressures in the absence of tidal positive pressure ventilation. CONCLUSIONS: Isolated phasic elevations of pulmonary artery pressure may cause less damage than those occurring during intermittent positive pressure mechanical ventilation, suggesting that cyclic changes in perivascular pressure surrounding extra-alveolar vessels may be important in the genesis of ventilator-induced lung injury.