OBJECTIVE: To evaluate aspects of reliability and validity of the Juvenile Arthritis Functional Status Index (JASI). We developed this questionnaire to assess a variety of daily living and functional mobility tasks in children with juvenile rheumatoid arthritis (JRA). JASI part I consists of 100 functional items divided into 5 activity categories (self-care, domestic, mobility, school, extracurricular). A 7 point degree of difficulty rating scale is used for responses. JASI Part II is a priority function section within which the child identifies and scores important activities for improvement. METHODS: The JASI was administered to 30 children with JRA 8 to 19 years of age at baseline, 3 weeks, and 3 months to determine test-retest reliability. Various rheumatology measures were used to evaluate the JASI construct validity. RESULTS: Reliability of JASI Part I was excellent (intraclass correlation coefficients > or = 0.95) for both retest intervals. Reliability was lower for respondents with mild disease than those with polyarticular disease. JASI Part I scores correlated strongly with rheumatology measures including joint count, grip strength, hip synovitis, timed walk and run. ACR functional class, and range of motion (r > 0.50), indicating construct validity. There was fair to good agreement between scores from the child's self-report and those from therapist observation of the child's abilities on the same activities. This provides evidence that children as young as 8 years of age can realistically report on function on their own. In JASI Part II, respondents identified a mean of 5.9 functional items in the baseline priority activity list. Test-retest reliability for JASI Part II was fair (kappa = 0.57). Further work is needed to determine the most reliable and sensitive response sequence for evaluating change in priority functions with JASI Part II. CONCLUSION: JASI Part I is a reliable and valid functional measure for school age children and adolescents with JRA. Research is under way to evaluate responsiveness to change. Comparisons will be made with shorter JRA functional measures to determine differences in sensitivity.