Background: To our knowledge, no data have been reported on the effects of exercise training (EXT) on central respiratory motor output or neuromuscular coupling (NMC) of the ventilatory pump, and their-potential association with exertional dyspnea. Accurate assessment-of these important clinical outcomes is integral to effective management of breathlessness of patients with COPD. Material and methods: Twenty consecutive patients with stable. moderate-to-severe COPD were tested at 6-week intervals at baseline, after a nonintervention control period (pre-EXT), and after EXT. Patients entered an outpatient pulmonary, rehabilitation program involving regular exercise on a bicycle. incremental symptom-limited exercise testing (1-min increments of 10 W) was performed on an electronically braked cycle ergometer. Oxygen uptake (Vo(2)), carbon dioxide output (VCO2), minute ventilation (VE), time, and volume components of the respiratory cycle and, in six patients, esophageal pressure swings (Pessw), both as actual values and as percentage of maximal (most negative in sign) esophageal pressure during sniff maneuver (Pessn), were measured continuously over the runs. Exertional dyspnea and leg effort were evaluated by administering a Borg scale. Results: Measurements at baseline and pre-EXT were similar. Significant increase in exercise capacity was found in response, to EXT: (1) peak work rate (WR), Vo(2), Vco(2), VE, tidal volume (VT), and heart rate increased, while peak exertional dyspnea and leg effort did not ;significantly change; (2) exertional dyspnea/Vo(2) and exertional dyspnea/Vco(2) decreased while VE/Vo(2) and VE/Vco(2) remained unchanged. The slope of both exertional, dyspnea and leg effort relative to VE fell significantly after EXT; (3) at standardized WR, VE, and Vco(2) exertional dyspnea and leg effort decreased while inspiratory capacity (IQ increased. Decrease in VE was accomplished primarily, by decrease in respiratory. rate (RR) and increase in both inspiratory time (TI) and expiratory time; VT slightly increased, while inspiratory drive (VT/TI) find duty cycle (TI/total time of the respiratory cycle) remained unchanged. The decrease in Pessw and the increase in VT were associated with lower exertional dyspnea after EXT, (4) at standardized VE, VT, RR, and IC, Pessw and Pessw(%Pessn)/VT remained unchanged while exertional dyspnea and leg effort decreased with EXT. Conclusion: In conclusion, increases in NMC, aerobic capacity, and tolerance to dyspnogenic stimuli and possibly breathing retraining are likely to contribute to the relief of both exertional dyspnea and leg effort after EXT.