The validity and reliability of self‐administered joint counts are reported in a group of32 rheumatoid arthritis patients being followed at a university‐based practice located in theSoutheast region of the United States, serving low to middle income urban and rural patients. Adequate inter‐rater reliability among the patients' and the research assistant's joint counts was obtained for upper (r = 0.74), lower (r = 0.96), and upper and lower extremities (r = 0.89). Convergent validity correlations for pain, helplessness, and the Joint Alignment and Motion scale were found to be adequate. We conclude that rheumatoid arthritis patients can reliably assess their joint counts. Self joint counts along with other validated self‐reportsof health status may be applicable to busy outpatient settings, as well as in clinical research. Copyright © 1993 WILEY‐VCH Verlag GmbH & Co. KGaA