The resting mouth occlusion pressure 0.1 s after onset of inspiration (P0.1) and minute ventilation (V'E) and their response to CO2 in patients with chronic obstructive pulmonary disease (COPD) remain controversial. The ventilatory drive and the factors that predict resting arterial CO2 tension (Pa,CO2) were studied in 19 eucapnic and 14 hypercapnic severe COPD patients, and 20 controls. The CO2 response was evaluated by the Read technique. The V'E, and P0.1 as a function of end-tidal CO2 tension (Pet,CO2) was used to study the ventilatory (deltaV'E/deltaPet,CO2) and P0.1 response (deltaP0.1/deltaPet,CO2). In the patients, respiratory muscle function and pleural occlusion pressure 0.1 s after onset of inspiration (Ppl,0.1) were evaluated with simultaneous measurement of pleural (Ppl) and gastric (Pga) pressures. Hypercapnic patients had lower forced vital capacity (FVC), forced expiratory volume in one second (FEV1), and arterial O2 tension (Pa,O2). Resting P0.1 was higher in patients than in controls, whereas deltaP0.1/deltaPet,CO2 was similar in the three groups. There was no difference in resting P0.1 (3.6+/-2.0 versus 4.3+/-2.8 kPa (2.7+/-1.5 versus 3.2+/-2.1 cmH2O), p=0.2) and Ppl,0.1 (1.4+/-2.3 versus 5.2+/-3.3 kPa (4.08+/-1.7 versus 3.9+/-2.5 cmH2O), p=0.22) between eucapnic and hypercapnic COPD, whereas deltaV'E/deltaPet,CO2 was lower in the hypercapnic group (0.29+/-0.24 versus 0.66+/-0.5 L x min(-1) x kPa, p<0.001). By logistic regression only FEVI and increased diaphragmatic load, and not respiratory drive, predicted resting Pa,CO2. Irrespective of CO2 level, baseline central drive (represented by the mouth occlusion and pleural pressures) and CO2 response are preserved in most patients with severe chronic obstructive pulmonary disease. Effective ventilation is inadequate in the more severely obstructed patients and this results in hypercapnia. Neuroventilatory coupling failure is an attractive explanation for chronic hypercapnia in these patients.