A group of 120 patients with rheumatoid arthritis or osteoarthritis volunteered to be subjects for this study of aerobic versus nonaerobic exercise. Patients were stratified by diagnosis and randomized into an exercise program of aerobic walking, aerobic aquatics, or non-aerobic range of motion (controls). The retention rate for the 12-week program was 83%. Exercise tolerance, disease-related measures, and self-reported health status were assessed. The aquatics and walking exercise groups showed significant improvement over the control group in aerobic capacity, 50-foot walking time, depression, anxiety, and physical activity after the 12-week exercise program. There were no significant between-group differences in the change scores for flexibility, number of clinically active joints, duration of morning stiffness, or grip strength. Our findings document the feasibility and efficacy of conditioning exercise for people who have rheumatoid arthritis or osteoarthritis.