Bilevel pressure ventilators are being used increasingly to provide noninvasive ventilatory support in the management of obstructive sleep apnea, chronic ventilatory failure, and acute respiratory failure. However, the ability of these ventilators to respond to inspiratory demand without imposing expiratory loads has not been evaluated extensively. We evaluated the performance of nine bilevel pressure ventilators in a lung model, as compared with the Nellcor Puritan-Bennett 7200ae adult critical care ventilator. All ventilators were set to provide pressure support ventilation (PSV) and positive end-expiratory pressure (PEEP) at a rate of 10 breaths/min with an inspiratory time of 1.0 s. Simulated pleural pressure, airway pressure, and flow at airway opening were continuously monitored. We studied the effects of three PSV levels (5, 10, and 15 cm H2O) with 5 cm H2O PEEP at two lung compliances (50 and 80 mL/cm H2O) and four peak inspiratory flow demands (20, 40, 60, and 80 L/min) on seven dependent variables: inspiratory delay time (D-I), inspiratory trigger pressure (P-I), inspiratory area percent (Area I%), expiratory delay time (D-E), supraplateau expiratory pressure change (P-E), expiratory area (Area E), and ventilator peak flow (VPF). Most ventilators performed as well as or significantly (p<0.05) better than the 7200ae in all studied variables. Compliance did not significantly affect ventilator performance. Increasing inspiratory flow demand significantly (p<0.05) increased D-I, P-I, P-E, and VPF and decreased Area I% with most ventilators. As ventilatory demand increased, D-E and Area E significantly (p<0.05) changed. With some units, D-E and Area E increased, while with others they decreased. Most bilevel pressure ventilators evaluated were able to respond to high ventilatory demands and outperformed the Nellcor Puritan-Bennett 7200ae ventilator.