Airway pressure, flow, and volume were measured before and after administration of aerosolized metaproterenol during controlled mechanical inflation and stepwise deflation of the relaxed respiratory system in 13 mechanically ventilated patients. An increase in passive expiratory flow at constant respiratory system recoil pressure was considered evidence of bronchodilatation. In 10 patients, at a respiratory system recoil pressure of 6 cm H2O (VP6), expiratory flow increased 21 to 500% above prebronchodilator level. In these 10 dynamically hyperinflated patients, an increase in VP6 was associated with a decrease in peak inspiratory pressure (Ppeak) (mean delta = -4.7 cm H2O) and a decrease in intrinsic positive end-expiratory pressure (Peepi) (mean delta = -2.4 cm H2O). The elastance of the respiratory system was not affected by metaproterenol, and the delta Peepi corresponded to a mean decrease in end-expiratory lung volume of 0.20 L. The results are consistent with predictions based on a single-compartment model. When mean expiratory flow is determined only by the tidal volume and expiratory time, a decrease in airway resistance results in a decrease in lung volume at which patients are ventilated. Therefore, the decrease in Ppeak is caused not only by a decrease in the resistive pressure cost but also by a decrease in the elastic pressure cost of inflating the respiratory system. It is emphasized that Ppeak and Peepi provide valuable information about bronchodilator-induced changes in lung function during controlled mechanical ventilation.