We investigated the mechanisms responsible for oxygen-induced hypercarbia in ventilator-dependent patients with advanced chronic obstructive pulmonary disease (COPD). To quantitate the effects of oxygen (O2) on respiratory drive, we determined the CO2 recruitment threshold (PCO2 RT) in 10 mechanically ventilated patients under normoxic (PaO2 = 67 +/- 7 mm Hg) and hyperoxic (PaO2 = 370 +/- 67 mm Hg) conditions. PCO2 RT is a measure of the CO2 responsiveness of the mechanically unloaded respiratory system and, as such, is independent of mechanical impedance and respiratory muscle strength. After O2 supplementation, PCO2 RT increased from 42 +/- 6 to 45 +/- 6 mm Hg (p less than or equal to 0.05), indicating a suppression of so-called hypoxic respiratory drive. The effect of hyperoxia on the dead space to tidal volume ratio (VD/VT) and CO2 elimination (VCO2) was studied in 6 patients. Measurements were made at identical ventilator settings, thus eliminating breathing pattern- and respiratory work-related effects on these variables. VD/VT rose from 0.49 +/- 0.09 to 0.55 +/- 0.06 (p less than or equal to 0.05), but VCO2 remained constant at 0.21 L/min. We discuss why measuring O2-induced changes in minute ventilation, VCO2, PaO2, and VD/VT in spontaneously breathing patients is insufficient to distinguish between gas exchange- and respiratory drive-related mechanisms for hypercarbia. Based on the O2-induced increase in PCO2 RT, we conclude that so-called suppression of hypoxic drive plays an important role in the pathogenesis of this disorder.