Rehabilitation programs for patients with chronic obstructive pulmonary disease (COPD) are common in clinical care throughout the Western world. However, the efficacy of these programs has not been evaluated in randomized experimental studies that have followed patients longer than one year or with outcome measures of important quality of life and psychosocial variables. We have conducted an experimental trial in which patients have been randomly assigned to comprehensive rehabilitation or to an education control group. The 119 patients participating in the program were evaluated prior to treatment, after two months, six months, one year, eighteen months, and two years. Outcome measures included both psychosocial and physiological variables. The psychosocial measures included a quality of well-being scale, measures of self-efficacy expectations, depression, and social support. The physiological measures included: pulmonary function tests, maximum exercise tolerance with measures of expired and arterial blood gases, and endurance exercise performance. Also perceived symptom ratings were taken with each exercise test. Data from the first year of follow-up have been evaluated and there are some significant differences between the comprehensive rehabilitation and education patients. In this application, we propose to extend the follow-up for four additional years. Physiological variables will be evaluated every other year. The psychosocial measures will be taken each year. In addition, mortality will be assessed for all participants using the National Death Index. Using the quality of life outcome data and data obtained on costs, we will construct a policy model evaluating the cost/utility of the rehabilitation interventions.