We postulated that ventilatory assistance during exercise would improve cardiopulmonary function, relieve exertional symptoms, and increase exercise endurance (T(lim)) in patients with chronic congestive heart failure (CHF). After baseline pulmonary function tests, 12 stable patients with advanced CHF (ejection fraction, 24 +/- 3% [mean +/- SEM]) performed constant-load exercise tests at approximately 60% of their predicted maximal oxygen consumption (V O(2)max) while breathing each of control (1 cm H(2)O), continuous positive airway pressure optimized to the maximal tolerable level (CPAP = 4.8 +/- 0.2 cm H(2)O) or inspiratory pressure support (PS = 4.8 +/- 0.2 cm H(2)O), in randomized order. Measurements during exercise included cardioventilatory responses, esophageal pressure (Pes), and Borg ratings of dyspnea and leg discomfort (LD). At a standardized time near end-exercise, PS and CPAP reduced the work of breathing per minute by 39 +/- 8 and 25 +/- 4%, respectively (p < 0. 01). In response to PS: T(lim) increased by 2.8 +/- 0.8 min or 43 +/- 14% (p < 0.01); slopes of LD-time, V O(2)-time, V CO(2)-time, and tidal Pes-time decreased by 24 +/- 10, 20 +/- 11, 28 +/- 8, and 44 +/- 9%, respectively (p < 0.05); dyspnea and other cardioventilatory parameters did not change. CPAP did not significantly alter measured exercise responses. The increase in T(lim) was explained primarily by the decrease in LD- time slopes (r = -0.71, p < 0.001) which, in turn, correlated with the reductions in V O(2)-time (r = 0.61, p < 0.01) and tidal Pes-time (r = 0.52, p < 0.01). in conclusion, ventilatory muscle unloading with PS reduced exertional leg discomfort and increased exercise endurance in patients with stable advanced CHF.