Using data from 3 studies of patients with rheumatoid arthritis, we examined the extent to which responses to items in the Center for Epidemiological Studies Depression Scale (CES-D) are influenced by aspects of the disease process other than depression. Our findings suggested that 4 CES-D items (i.e., "I felt that everything I did was an effort," "I felt hopeful about the future," "My sleep was restless," and "I could not get going") may be influenced by aspects of the disease process and, thus, are not necessarily indicative of depression among persons with arthritis. The impact that these items have on the interpretation of CES-D scores was assessed in relation to 2 research issues: estimation of the prevalence and severity of depression in arthritis populations and identification of the determinants of depression among individuals with arthritis. Our results suggest that the original CES-D may overestimate the prevalence and severity of depression among patients with arthritis. The magnitude of this bias is modest, however. The results also suggest that in studies designed to identify the determinants of depression among individuals with arthritis, inclusion of the 4 items identified is unlikely to have any effect on study findings.