This study tests the efficacy of respiratory mechanical unloading as a mode of assisted mechanical ventilation in cats with an intact breathing-control system but severe pulmonary parenchymal injury. Twelve anaesthetized, intubated cats received multiple saline lung lavages so that their total respiratory system compliance decreased from 56.1+/-10.4 to 26.8+/-6.8 ml/kPa (p < 0.001) and their PaO2 fell to 12.38+/-4.71 kPa when 100% O2 was used as inspired gas. They were then exposed to three consecutive 15-min periods of CPAP of 0.5 kPa, respiratory unloading and again CPAP of 0.5 kPa. Unloading was applied with end-expiratory pressure of 0.5 kPa, elastic assistance of 0.03 kPa/ml and resistance compensation of 2.0 kPa/l/s. Arterial blood gases for the CPAP baselines did not differ significantly before and after unloading: pH 7.14+/-0.04 vs. 7.16+/-0.06; PaCO2 8.99+/-2.07 vs. 8.33+/-2.01 kPa; PaO2 12.4+/-4.7 vs. 13.3+/-7.6 kPa. Nor did the baselines differ in terms of tidal volume, respiratory rate and phrenic nerve activity. Unloading increased tidal volume substantially by about 50% and increased respiratory rate slightly, while inspiratory time remained unchanged. PaCO2 fell to 6.63+/-1.57 kPa and pH rose to 7.25+/-0.06. Phrenic nerve activity was significantly down-regulated in terms of total number of impulses and mean impulse frequency in the phrenic nerve burst. These results suggest that combined elastic and resistive unloading may be an effective means of assisted mechanical ventilation in severe respiratory failure of pulmonary parenchymal origin.