Rheumatoid arthritis (RA) is an aggressive and progressive condition that has an extraordinary impact on the individual and society in general (Brooks, 1997). For the individual patient, a shortened life expectancy has been documented, with standard mortality ratios predicted to be 1.3 to 3.0 in clinical studies (Wolfe, 1997). Quality of life is also affected. Work disability is markedly reduced in this disease, with estimates up to 50% of patients unable to work full time within 10 years from the onset of their illness (Yelin 1995). This research examines whether cognitive behavior therapy interventions that target responses to chronic stress, pain, and depression will reduce disease activity and improve the mental health of older adults with RA. The study combines field assessments, laboratory tests of stress reactivity and clinical evaluations of the person's condition in a longitudinal design. After pre-testing, 210 RA participants are randomly selected into one of three treatments: Cognitive-Behavior Therapy for Pain (CBT-P), Cognitive-Behavior Therapy for Depression (CBT-D) or Education Group only (EG), which serves as a control. Illness severity, depressive symptoms and interpersonal difficulties are expected to predict psychological and physiological stress responses. Stress reactivity measures are expected to predict the course of illness over time. Those receiving CBT-D are expected to show the greatest improvement in physical functioning, mental health and disease activity. Successful treatment is expected to alter stress responses, leading to better health and mental health over time.