Rheumatoid arthritis (RA) is a painful systemic disease and is believed to be exacerbated by stress. Relaxation and biofeedback strategies have demonstrated utility in alleviating both pain and stress-related symptomatology, and therefore were tested for efficacy with this disease in a two-phase study. First, 24 patients were taught a relaxation technique and then trained in either temperature elevation or reduction. Second, a group of 15 patients thus trained was compared with 8 others who received traditional physiotherapy modalities. Psychological tests, functional/physical evaluations, as well as measurements related to pain, sleep, and other activities were carried out. Results of the first study revealed significant and positive changes following treatment that were primarily related to pain, tension, and sleep patterns for both groups, but no differential effects were noted between temperature elevation or reduction conditions. This was attributed to both groups having maintained temperature above baseline during biofeedback training. The results of the second study consistently favored the relaxation and biofeedback over the physiotherapy group on the physical/functional indices. The psychological measures tended to remain constant throughout both studies, leading to the conclusion that the effectiveness of treatment was specific to physical functioning rather than to a psychological enhancement of well-being.