Hypercapnia in patients with pulmonary disease is believed to result from an interaction between mechanical lung impairment and intrinsic chemical respiratory drive. We tested this hypothesis in this study by examining the ventilatory (delta VE/delta PCO2) and occlusion pressure (delta P100/delta PCO2) responses to CO2 in 12 obese patients with no history of alveolar hypoventilation and correlating these with their ventilatory responses to abdominal surgery. Preoperatively the mean vital capacity (VC) was 78% +/- 6% standard error of the mean predicted, the delta VE/delta PCO2 was 1.56 +/- 0.26 L/min/torr, delta P100/delta PCO2 was 0.25 +/- 0.08 cm H2O/torr, the mean PaCO2 37.9 +/- 1.1 mm Hg, and mean PO2 77.6 +/- 3.7 mm Hg. Postoperatively the VC decreased to 56% +/- 6% of the preoperative value. PCO2 values at 24 hours increased in six patients, were unchanged in three, and decreased in three patients. However, over the entire spectrum of PCO2 change, both indexes of CO2 chemosensitivity correlated strongly with the postoperative change in PCO2 (r = -0.86 for delta VE/delta PCO2 and r = -0.66 for delta P100/delta PCO2). All six patients with a delta VE/delta PCO2 of 1.5 L/min/torr or less manifested postoperative increases in PCO2, while those with greater values did not (p = 0.005). In contrast, neither preoperative nor postoperative VC showed high correlations with postoperative PCO2 (r = -0.56 and -0.43, respectively). Thus ventilatory responses to CO2 predicted postoperative PCO2 at both ends of the spectrum; low responders hypoventilated while high responders hyperventilated. We conclude that in obese subjects, CO2 chemosensitivity plays a permissive role in determining the net ventilatory responses to situations that either mechanically load the respiratory system or modulate ventilation such as postoperative pain or analgesia.