Two instruments measuring handicap were evaluated and compared with clinical, laboratory and disability measures. Participants were 133 patients attending a rheumatology follow-up clinic in a district general hospital, of whom 102 were followed up after 3 months. Measurements included acute-phase response, early morning stiffness, pain, wellbeing, joint involvement (impairments), the Stanford Health Assessment Questionnaire (disability), the Disease Repercussion Profile and the London Handicap Scale (handicap). A substantial burden of disability and handicap was recorded. There were moderate correlations between impairments (0.4 < rho < 0.6), and moderate to strong correlations between disability and handicap measures (0.4 < rho < 0.8). Correlations between impairment and disability/handicap were weak (rho < 0.4). Mean changes in all variables over 3 months were small, and none was statistically significant. A comprehensive description of the impact of disease and treatment requires measurements to be made of impairments, disabilities and handicaps. The use of clinical and laboratory variables alone may be misleading.