It has been reported that, in diseased lungs, either partial liquid ventilation (PLV) or high-frequency oscillatory ventilation (HFOV) can improve oxygenation better and with less lung injury than conventional mechanical ventilation (CMV). This study was intended as a preclinical comparison between the effects of HFOV, PLV and CMV on gas exchange, lung mechanics and histology. Fifteen anesthetized newborn piglets, with respiratory insufficiency due to repeated saline lung lavage, were allocated to either a PLV, HFOV or CMV (n = 5 each) strategy, and treated for 4 h. Within 30 min of commencing therapy, PLV, HFOV, and CMV improved arterial PO2 (Pa,O2), alveoloarterial oxygen gradient (P(A-a),O2), oxygenation index (OI), venous admixture (va), and arterial PCO2 (Pa,CO2). After 4 h, oxygenation parameters (Pa,O2, P(A-a),O2, OI and venous admixture) were significantly better in the HFOV group than in the PLV group; the CMV group showed a higher Pa,O2 and lower OI than the PLV group. Gas exchange at the end of the experiment was not different from baseline in the HFOV and CMV groups. Lung histology and morphometry were performed after perfusion-fixation at endotracheal deflation pressure corresponding to mean airway pressure at the end of the experiment. Lung injury score and mean linear intercept were not different between the three treatment groups. We conclude that in this model, gas exchange improved significantly in all three ventilation strategies. Indices of oxygenation improved less during PLV. The saline lavage-induced acute lung injury model used as in this study, is less stable than previously thought. The final lung injury is not influenced by the ventilation strategy. We speculate that the impaired gas exchange during PLV is an expression of diffusion limitation and ventilation-perfusion mismatch in a recovering lung.