Abstract： Objective To investigate the impact of proportional assist ventilation (PAV) on tolerance and breathlessness in ventilated chronic obstructive pulmonary disease (COPD), a nd to describe the patient-ventilator interaction, hemodynamic state, breathing pattern and work of breath during PAV and pressure support ventilation (PSV) . Methods Ten intubated COPD patients on weaning from mechanical ventilation were studied . Elastance and resistance were measured by both the inspiratory-hold techniqu e during a brief period of volume control ventilation and runaway technique duri ng PAV. Each assistance level of PAV (80%, 60% and 40%) and PSV was selected ra ndomly. Patients' response, hemodynamics, blood gas and lung mechanics were mon itored. Results Tidal volume and respiratory rate didn't change in a consistent manner as the le vel of assist was decreased (P>0.05). With the level of assist increasing, peak inspiratory pressure was increasing significantly (P<0.05), while pat ients' work of breath had the tendency to decrease (P<0.05). A significant difference in the Borg Category Scale was observed between PAV and PSV (0.50[ 1.50] vs 0.75[2.00], P<0.05) at the same degree of respiratory muscle unloading. PaCO2 was significantly higher on PAV (54[23]mmHg) than on PSV (48[23]mmHg) (P<0.05). Peak inspiratory pressure on PAV was sign ificantly lower than on PSV (16±4cmH2O vs 21±3cmH2O, respectively , P<0.05). Hemodynamics and oxygenation remained unchanged.Conclusions PAV is a feasible method for supporting ventilator-dependent patients and was w ell tolerated. It can improve the breathing pattern and reduce inspiratory effo rt. At the same degree of respiratory muscle unloading, PAV can be implemented at much lower peak inspiratory pressure than PSV. It can also apply proportiona l pressure support according to the patients' ventilatory demand.