The applicability of high-intensity training and the possibility of inducing physiologic adaptation to training are still uncertain in patients with severe chronic obstructive pulmonary disease (COPD). The purposes of this study were to evaluate the proportion of patients with moderate to severe COPD in whom high-intensity exercise training (30-min exercise session at 80% of baseline maximal power output [Wmax]) is feasible, and the response to training in these patients. We also sought to evaluate the possible influence of disease severity on the training intensity achieved and on the development of physiologic adaptation following endurance training. Forty-two patients with COPD (age = 66 +/- 7 yr, FEV1 = 38 +/- 13% predicted, [mean +/- SD]) were evaluated at baseline and after a 12-wk endurance training program. Each evaluation included a stepwise exercise test on an ergocycle up to the individual maximal capacity during which minute ventilation (VE), oxygen consumption (VO2), carbon dioxide production (VCO2), and arterial lactic acid concentrations were measured. The training consisted of 25 to 30-min exercise sessions on a calibrated ergocycle three times a week, with a target training intensity at 80% of Wmax. The training intensity was adjusted with the objective of reaching the target intensity, but also to ensure that the cycling exercise could be maintained for the specified duration. The training intensity sustained for the duration of each exercise session averaged 24.5 +/- 12.6, 51.7 +/- 17.4, 63.8 +/- 22.4, and 60.4 +/- 22.7% of Wmax at Weeks 2, 4, 10, and 12, respectively. High-intensity training was achieved in zero, three, five, and five patients at Weeks 2, 4, 10, and 12, respectively. A significant increase in VO2max and Wmax occurred with training (p < 0.0002). This improvement in exercise capacity was accompanied by a 6% and 17% reduction in VE and in arterial lactic acid concentration for a given work rate, respectively (p < 0.0001), suggesting that physiologic adaptation to training occurred. The intensity of training achieved, in % Wmax, was not influenced by the initial VO2max, age, or FEV1. The effects of training were compared in patients with an FEV1 > or = 40% or < 40% predicted. Percent changes in VO2max, Wmax, and VE, were significant and of similar magnitude for both groups, whereas the decrease in arterial lactic acid for a given work rate reached statistical significance only in those patients with an FEV1 > or = 40% predicted. We conclude that although most patients were unable to achieve high-intensity training as defined in this study, significant improvement in their exercise capacity was obtained and physiologic adaptation to endurance training occurred. The training intensity expressed as a percent of the individual maximum exercise capacity, and the relative effectiveness of training, were not influenced by the severity of airflow obstruction.