Cognitive-behavioral therapy (CBT) interventions are often used as adjunct treatments with standard medical care to help patients with chronic illnesses better manage their pain and distress or improve function. We review the primary assumptions and the four essential components that underlie all CBT interventions. We then examine the outcomes produced by CBT interventions for patients with two chronic illnesses for which the source of pain is well understood (rheumatoid arthritis and osteoarthritis of the knee) and three disorders characterized by "medically unexplained" pain and other symptoms (fibromyalgia, chronic fatigue syndrome, irritable bowel syndrome). We employed American Psychological Association criteria for empirically validated treatments to evaluate the CBT intervention outcomes. Using these criteria, we conclude that CBT is a well-established adjunct treatment for patients with rheumatoid arthritis and is probably efficacious for patients with knee osteoarthritis, chronic fatigue syndrome, and irritable bowel syndrome. However, CBT interventions must be considered as an experimental therapy for patients with fibromyalgia. The negative findings associated with CBT for fibromyalgia may be due to several factors such as high levels of psychiatric morbidity or abnormal pain sensitivity among the patients. It is necessary for investigators to devote effort to identifying the factors that account for poor CBT outcomes and to modify current CBT protocols to address these factors. © 2003 Elsevier Inc. All rights reserved.