In order to evaluate the efficacy of ventilation by high frequency transthoracic oscillation in obstructive lung disease, we studied 10 anesthetized and paralyzed rabbits during experimental bronchoconstriction. In order to generate body surface oscillation, the animals were placed in a body-box connected to a piston pump. Two frequencies were successively applied in random order: 1 Hz, representing "conventional" ventilation, and 5 Hz for high frequency ventilation. During the control period, the magnitudes of the oscillation were adjusted such that the tidal volumes thus produced yielded normocapnia (PaCO2 = 40 +/- 2 mmHg) for at least 3 min at each frequency. The same tidal volumes were then used after a bronchoconstriction induced by aerosolized carbachol. We obtained a severe and stable bronchoconstriction and oscillated the animals at 1 and 5 Hz at identical tidal volumes as those used during the control period. Identical at the 2 frequencies of oscillation during the control period, the rate of CO2 elimination with bronchoconstriction was significantly higher (p less than 0.05) at 5 Hz (15.9 +/- 3.3 ml/min) than at 1 Hz (14.2 +/- 3.9 ml/min); PaCO2 was also significantly lower at 5 Hz than at 1 Hz. Mean arterial PO2 was significantly higher (p less than 0.01) at 5 Hz (58 +/- 4 mmHg) than at 1 Hz (46 +/- 12 mmHg) during bronchoconstriction. Mean alveolararterial O2 pressure difference was significantly lower (p less than 0.01) at 5 Hz (49 +/- 13 mmHg) than at 1 Hz (59 +/- 11 mmHg). Only a minor increase in lung volume was found during high frequency ventilation, and no significant hemodynamic changes occurred.(ABSTRACT TRUNCATED AT 250 WORDS)