We evaluated the relationship of clinical characteristics, pulmonary function, and exercise test data to the degree of improvement in the 12-minute walking distance (12MD) in 50 ambulatory outpatients completing a six-week pulmonary rehabilitation program. The 12MD increased by 27.7 +/- 32.5 percent, or 462 +/- 427 ft, by the end of the program. There were no significant relationships between improvement in the 12MD and age, sex, oxygen requirement, arterial blood gas levels, and pulmonary function; however, patients with a greater ventilatory reserve (1-[VEmax/MVV] x 100) had more improvement in their 12MD, both with respect to distance and percentage of increase over baseline. Additionally, patients with a lower peak oxygen consumption (VO2) and peak oxygen pulse (O2P) showed greater percentage of improvement in their 12MD. The magnitude of the initial 12MD was inversely related to its improvement, both with regard to distance (r = -0.43; r2 = 0.18; p less than 0.003) and percentage of increase (r = -0.71; r2 = 0.51; p less than 0.0001). Using stepwise regression, the combination of smaller initial 12MD and greater FEV1 was significantly predictive of improvement in the 12MD. Patients with poor performance on either a 12MD or maximal exercise test are not necessarily poor candidates for a pulmonary rehabilitation program.