This study investigates the impact of proportional assist ventilation (PAV), a new mode of partial ventilatory support, on exercise tolerance and breathlessness in severe hypercapnic chronic obstructive pulmonary disease (COPD) patients. We also examined the effects of continuous positive airway pressure (CPAP) and pressure support ventilation (PSV). On two consecutive days, 15 stable hypercapnic COPD patients underwent four endurance tests on a cycle ergometer at 80% of their maximal workrate, receiving, via a nasal mask in random order, either: 1) sham ventilation (CPAP: 1 cmH2O); 2) CPAP (6 cmH2O); 3) PSV (inspiratory pressure support: 12-16 cmH2O; expiratory positive airway pressure (EPAP): 1 cmH2O); or 4) PAV (8.6+/-3.6 cmH2O x L(-1) and 3+/-1.3 cmH2O x L(-1)x s(-1) of volume and flow assistance, respectively plus EPAP: 1 cmH2O). Oxygen supply was standardized to maintain an arterial oxygen saturation (Sa,O2) of 92-93%. Breathing pattern and minute ventilation (by respiratory inductive plethysmography), pulse oximetry, end tidal partial pressure of CO2, cardiac frequency and sensations of dyspnoea and leg discomfort (by Borg scale) were monitored. In comparison to sham ventilation, PAV, PSV and CPAP were able to increase the endurance time (from 7.2+/-4.4 to 12+/-5.6, 10+/-5.2 and 9.6+/-4.6 min, respectively) and to reduce dyspnoea and oxygen flow to the nasal mask. However, the greatest improvement was observed with PAV. We conclude that PAV delivered by nasal mask can im-prove exercise tolerance and dyspnoea in stable hypercapnic COPD patients and hence this mode of ventilatory support may be useful in respiratory rehabilitation programmes.