Controversy exists as to whether high-frequency oscillatory ventilation can be used on babies and small laboratory animals only, or whether high-frequency oscillatory ventilation can also be efficient in the adult patient and large (> 65 kg body weight) laboratory animals. Moreover, controversy exists as to whether limitations in high-frequency oscillation efficiency are caused by the size and shape of the bronchial system, by the lack of low impedant intersegmental gas flow in lung parenchyma, or by inappropriate high-frequency ventilators and ancillary hardware. Therefore, our objective in this study using the adult pig as a model of the adult patient was to test whether the adult airway system is suited to the use of high-frequency oscillatory ventilation or whether there are geometrical, structural, or functional limitations to efficient ventilation by high-frequency oscillation. Prospective, controlled, randomized comparison over 8 to 16 hrs of ventilatory management. Experimental thoracovascular surgery laboratory in a university hospital. Fifteen adult, female, house swine (weight 90 to 140 kg). We evaluated the ventilatory effect of a wide range of oscillation frequencies (10-15 to 35-45 Hz), tidal volumes (0.5 to 2.2 mL/kg), and bias flow volumes (10 to 70 L/min) at a mean airway pressure of 12 +/- 1 cm H2O in anesthetized and relaxed pigs who did not have lung injury. Arterial blood gases are mainly dependent on tidal volume, frequency, and mean airway pressure. A threshold bias flow volume of 35 +/- 5 L/min is required to prevent CO2 rebreathing. In the group of lightweight animals (65 to 99 kg), the most efficient frequency band for CO2 elimination was approximately 25 Hz. The most efficient frequency band for arterial oxygenation was found to vary between individuals more than the most efficient frequency band for CO2 elimination. In the group of heavy animals (100 to 140 kg), no most efficient mean frequency could be assessed, probably because the excitation system was limited. We confirmed that tidal volume on its own had an effect on CO2 elimination ("tidal-volume effect"), although CO2 elimination was mainly determined by the product of tidal volume and oscillation frequency (oscillated minute volume), at least up to a critical frequency. Beyond that frequency, CO2 elimination could not be enhanced. The most efficient mean airway pressure in unimpaired lungs was assessed at 12 +/- 1 cm H2O. Adult pigs with a body weight in the range of the weight of clinical adult patients can be ventilated by high-frequency oscillation at tidal volumes smaller than, equal to, or slightly more than anatomical deadspace. The most efficient frequency for gas exchange varied between individuals. Tidal volume had an enhancing effect on CO2 elimination. The frequency dependency of PaO2 may have been related to a frequency-dependent structural remodeling of the airway system, which occurred even though the mean airway pressure was kept constant. These results demonstrate that failure of adequate ventilation by high-frequency oscillation is caused by a) CO2 rebreathing, b) the avoidance of an appropriate alveolar recruitment strategy, and c) an underpowered, high-frequency ventilatory system (oscillator) that is unable to deliver appropriate pressure oscillations. These limitations led to insufficient CO2 elimination and/or inadequate arterial oxygenation.