BACKGROUND: Spontaneous breathing (SB) activity may improve gas exchange during mechanical ventilation mainly by the recruitment of previously collapsed regions. Pressure support ventilation (PSV) and biphasic positive airway pressure (BIPAP) are frequently used modes of SB, but little is known about the mechanisms of improvement of lung function during these modes of assisted mechanical ventilation. We evaluated the mechanisms behind the improvement of gas exchange with PSV and BIPAP. METHODS: Five pigs (25-29.3 kg) were mechanically ventilated in supine position, and acute lung injury (ALI) was induced by surfactant depletion. After stabilization, BIPAP was initiated with lower continuous positive airway pressure equal to 5 cmH₂O and the higher continuous positive airway pressure titrated to achieve a tidal volume between 6 and 8 mL/kg. The depth of anesthesia was reduced, and when SB represented ≥20% of total minute ventilation, PSV and BIPAP + SB were each performed for 1 h (random sequence). Whole chest helical computed tomography was performed during end-expiratory pauses and functional variables were obtained. Pulmonary blood flow (PBF) was marked with IV administered fluorescent microspheres, and spatial cluster analysis was used to determine the effects of each ventilatory mode on the distribution of PBF. RESULTS: ALI led to impairment of lung function and increase of poorly and nonaerated areas in dependent lung regions (P < 0.05). PSV and BIPAP + SB similarly improved oxygenation and reduced venous admixture compared with controlled mechanical ventilation (P < 0.05). Despite that, a significant increase of nonaerated areas in dependent regions with a concomitant decrease of normally aerated areas was observed during SB. In five of six lung clusters, redistribution of PBF from dependent to nondependent, better aerated lung regions were observed during PSV and BIPAP + SB. CONCLUSIONS: In this model of ALI, the improvements of oxygenation and venous admixture obtained during assisted mechanical ventilation with PSV and BIPAP + SB were explained by the redistribution of PBF toward nondependent lung regions rather than recruitment of dependent zones.