This study examined the validity and reliability of full and short versions of the Arthritis Impact Measurement Scales (AIMS). One hundred fifty‐five patients with Rheumatoid Arthritis followed at a University Hospital Rheumatology Clinic completed the full AIMS at baseline, 6 months, 12 months, and 18 months. After reducing the 45‐item AIMS to 22, alpha reliabilities and test‐retest correlations showed that, with the exception of test‐retest correlations for mobility at 6 months and for pain at 12 and 18 months, the full and short scales were comparably reliable. Convergent validity correlations with theoretically related scales were also comparable. However, some of the short scales did not detect the same differences over time that the full scales did. Specifically, the short mobility, pain, anxiety, and depression scales were not as sensitive to change as the full scales. Except for these four scales, the short version appears to be a viable alternative for use by health professionals and researchers. Copyright © 1991 American College of Rheumatology