We tested the hypothesis that intermittent ventilatory assistance in patients with severe chronic obstructive pulmonary disease (COPD) improves pulmonary function and exercise capacity. Twenty stable patients with severe COPD were recruited from outpatient pulmonary clinics and were randomized to use a poncho wrap, negative-pressure ventilator or to receive standard care. After 6 months, the patients receiving standard care were switched over to the ventilator and vice versa, and follow-up was continued for an additional 6 months. After 3 to 6 months of ventilator use, we observed no clinically significant improvements in FEV1, FVC, blood gas determinations, maximal inspiratory and expiratory pressures, and exercise duration. However, 11 of our patients dropped out of the study because of an inability to tolerate the ventilator, and all but one of the nine who completed the study expressed dissatisfaction with it, using it for less time (4.1 h/day) than we recommended. Musculoskeletal pain and inconvenience were the most frequently voiced complaints. Because we did not document that ventilator use actually rested the respiratory muscles in our patients and because duration of ventilator use may have been too brief, we cannot conclude that intermittent rest of respiratory muscles in patients with severe COPD fails to bring about improvement. On the other hand, our results demonstrate that the poncho wrap ventilator is poorly tolerated by patients with severe COPD in a typical outpatient setting. We suggest that future trials seek to utilize better tolerated ventilatory assist devices.