We investigated the effect of ultrasonic nebulization versus instillation of exogenous surfactant on gas exchange abnormalities provoked by detergent inhalation in perfused rabbit lungs. Ventilation-perfusion (VA/Q) distribution was assessed by the multiple inert gas elimination technique. For nebulization of natural bovine surfactant (Alveofact), an ultrasonic device was placed in line with the inspiratory gas flow tubing, manufacturing particles with a mass median aerodynamic diameter of approximately 4.5 microM and high aerosol concentration. In vitro studies demonstrated biochemical and biophysical integrity of postnebulization surfactant. Lung aerosol deposition was monitored by a laser-photometric technique. In lungs with sham inhalation of saline, tracheal instillation of surfactant (approximately 11 mg/kg body weight, infused over 50 min) provoked substantial VA/Q mismatch and limited shunt flow, whereas lung surfactant deposition by ultrasonic nebulization (approximately 7 to 9 mg/kg body weight; nebulization time, 50 min) did not interfere with physiologic gas exchange. Tween 20 inhalation provoked severe VA/Q mismatch with predominant shunt-flow (approximately 21%). This was not reversed by "rescue" application of instilled surfactant, but largely reversed by nebulized surfactant (shunt reduced to 5.5%; p < 0.01). Analysis of postaerosol lavage fluid demonstrated partial reconstitution of surface activity by nebulized surfactant. We conclude that ultrasonic nebulization may be employed for efficient delivery of functionally intact natural surfactant to the distal bronchoalveolar space. This approach effects rapid improvement of gas exchange in a model of acute homogeneous lung injury.