Background and Purpose. Pain and physical function are core outcome measures for people with osteoarthritis, and self-report questionnaires have been the preferred assessment method. There is evidence suggestingthat self-reports of physical function represent what people experience when performing activities rather than their ability to perform activities. The purpose of this study was to examine the factorial validity of performance-specific assessments of pain and function. Subjects. The sample consisted of 177 participants who had osteoarthritis of the hip (n=81) or knee(n=96) and who were awaiting total joint arthroplasty. Methods. Through a cross-sectional design, participants performed 4 performance activities (self-paced walk test, stair test, Timed “Up & Go” Test, and Six-Minute Walk Test). Outcomes were time or distance (function) and pain ratings obtained immediately after each activity. The authors conceptualized 2 correlated factors, with pain items loading uniquely on 1 factor and functional items loading on the second factor, and uncorrelated error terms. Confirmatory factor analysis was applied. Results. Initial analysis yielded results consistent with the conceptualized model in this study with the exception of a nonzero correlation between the stair pain and function error terms. Dropping the stair test provided results consistent with the conceptualized model. Discussion and Conclusion. Given the limitations of self-report alone as a method of obtaining reasonably distinct assessments of pain and function, the extent to which performance-specific assessments could accomplish this goal was examined in this study. It was found that collectively the walk test, Timed “Up & Go” Test, and Six-Minute Walk Test yielded 2 factors consistent with the health concepts of pain and function. The authors believe that the application of these tests may provide clinicians and clinical researchers with more distinct impressions of pain and function that complement information from self-report measures.