The pulmonary function test (PFT) alone may be inadequate for predicting work-related exercise capacity in patients who file workers' compensation claims for respiratory limitation and compensation. Two hundred sixteen ambulatory patients with chronic obstructive pulmonary disease (forced expiratory volume in 1 second = 54.1 +/- 16.8% predicted) were administered the PFT and cardiopulmonary exercise test, and the results were analyzed by categorical statistical comparison, based on standard medical impairment classifications. Sixty-five patients (30.1%) were similarly classified by the two methods. Of the remaining patients, 132 (61.1%) were found to be less impaired according to the cardiopulmonary exercise test than according to the PFT, and 19 (8.8%) were more impaired according to the PFT. The results favor the use of the cardiopulmonary exercise test for the routine evaluation of respiratory impairment in patients with chronic obstructive pulmonary disease, particularly for patients with mild or moderate impairment revealed by the PFT. The large discrepancy between the two procedures emphasizes the need for a novel approach.