Rheumatoid arthritis (RA) clinical trials often include at least 10 measures of disease activity (e.g., tender and swollen joint counts, erythrocyte sedimentation rate, patient pain, grip strength, etc.), and each trial contains different measures. Using dissimilar disease activity (outcome) measures to assess therapies makes it impossible to judge therapies against a common standard. Furthermore, multiple statistical testing is performed, and outcome measures are often used which are insensitive to change. Ultimately, to compare different RA therapies, the same outcomes should be measured in all trials. In the first stage of an effort to develop a core set of outcome measures to be used in all RA trials, we present an approach to the selection of the measures. This approach is based on a critical evaluation of whether commonly used outcome measures have construct, face, content, criterion, and discriminant validity. In addition, our approach includes testing whether these measures are redundant. Also, selected evidence on the validity of currently used outcome measures is reviewed.