Submaximal constant work rate exercise tests are often used to measure the efficacy of an exercise intervention, but the reliability of these tests in patients with chronic obstructive pulmonary disease (COPD) has not been established. OBJECTIVE To examine the reproducibility of submaximal exercise responses of COPD patients compared with those of healthy elderly subjects. METHODS Sixteen COPD patients and 15 healthy subjects performed four weekly submaximal exercise tests against two different constant work rates: 20 W and 50% of the peak work rate (PWR). Spirometry was performed before each exercise test. COPD patients and healthy subjects were: age 69 +/- 5 and 65 +/- 5 yr, body mass index 26.4 +/- 3.9 and 26.7 +/- 3.0 kg x m(-2), respectively. RESULTS Both groups had no change in minute ventilation (V(E)), oxygen uptake (VO2), breathlessness (RPB), and leg fatigue (RPLF) for either work rate over repeated measures (P > 0.05). At 50% PWR test-retest reliability coefficients for V(E) and VO2 ranged from r = 0.88 to r = 0.96 for COPD patients and from r = 0.72 to r = 0.97 for healthy subjects; for RPB and RPLF test-retest reliability ranged from r = 0.76 to r = 0.89 for COPD patients and from r = 0.70 to r = 0.91 for healthy subjects. Intrasubject mean absolute differences for repeated measures of V(E), VO2, RPB, or RPLF were low and there were no group differences (P > 0.05). Percent error for V(E) and VO2 ranged from 6 +/- 3 to 9 +/- 7%, and for RPB and RPLF ranged from 19 +/- 18 to 68 +/- 65% for both groups at each work rate. CONCLUSIONS Submaximal exercise responses were reliable over a 1-month period, and measures of the physiological and psychophysical responses of COPD patients were as reliable as those of healthy subjects.