If chronic hypercapnia in patients with severe COPD occurs as a consequence of respiratory muscle (RM) weakness or fatigue, we would expect that ventilatory muscle recruitment (VMR) and exercise performance in stable hypercapnic patients would differ from those in eucapnic patients. We evaluated exercise performance and RM function at rest and during exercise in 19 eucapnic (PCO(2) 40 +/- 3 mm Hg), and 13 hypercapnic (PCO(2) 52 +/- 10 mm Hg) patients with severe COPD. A metabolic cart was used to determine V E, V O(2), V CO(2), and HR. Gastric (Pg) and esophageal (Ppl) balloons were used to measure Pg, Ppl, and Pdi. Ventilatory muscle recruitment pattern (VMR) was partitioned using end-inspiratory and end-expiratory Pg and Ppl. Hypercapnic patients had lower FEV(1) (0.60 +/- 0.24 versus 0.95 +/- 0.31 L, p < 0.001), MVV (28 +/- 11 versus 41 +/- 13 L, p < 0.001), resting PO(2) (61 +/- 11 versus 70 +/- 11 mm Hg, p < 0.001), peak PO(2) (60 +/- 20 versus 75 +/- 22 mm Hg, p < 0.005), and V E(max) (24 +/- 10 versus 32 +/- 12 L/min, p < 0.001). Patients in both groups had similar FRC (5.7 +/- 1.6 versus 5.0 +/- 1.5 L), V O(2)max (0.58 +/- 0.30 versus 0.76 +/- 0.32 L/min), Watts (45 +/- 48 versus 71 +/- 59), V E/MVV (88 +/- 33 versus 79 +/- 14), and HRmax (117 +/- 17 versus 128 +/- 18 beats/min). PI(max) (67 +/- 28 versus 65 +/- 32 cm H(2)O) and PE(max) (98 +/- 34 versus 96 +/- 40 cm H(2)O) were also similar in both groups. VMR (DeltaPg/DeltaPpl) at rest (-0.28 +/- 0.51 versus 0 +/- 0.35) and during exercise (0.4 +/- 0.2 versus 0.39 +/- 0.15) was equally affected in both groups. We conclude that exercise capacity and ventilatory muscle recruitment are similarly impaired in eucapnic and hypercapnic patients with severe COPD. These findings make inability of the lung to increase ventilation and not respiratory muscle dysfunction a more attractive explanation for CO(2) retention in stable hypercapnic patients.