MAC pain score1 1]/[daily activity count]) 3 1,000,000. We focus on pain in relation to activity because patients with joint disease commonly modify their activity in order to avoid pain (3). By incorporating both pain and activity into one score, we increase the dispersion of scores among those who do not have pain symptoms. In creating the PAKS score, we wanted to provide a single score that could reflect the pain level given a particular amount of activity experienced. Although it might be effective to fit a model for predicting the K/L grade that includes pain, activity, the interaction of pain and activity, and other risk factors, it was not our purpose to look for all predictors of K/L grade. Instead, we sought to determine whether a symptom score that includes an assessment of both pain and activity would be better than a pain score alone for discriminating between K/L grades. Our study demonstrated that PAKS scores discriminate radiographic OA severity better than WOMAC pain scores alone. A second concern was that PAKS scores require an assumption that a patient’s activity is a consequence of the severity of OA in the index knee. In fact, use of the PAKS score does not include this assumption. Although we believe that the severity of knee pain tends to increase with increased activity, this does not require that we ascribe the motivation for activity to the severity of OA. Irrespective of whether the physical activity level is driven by knee pain or other social, economic, cultural, environmental, psychological, or medical factors, our contention is that the salient information needed to interpret the severity of a patient’s knee pain is the amount of activity to which a knee is exposed. The PAKS score is intended to be an activity level–adjusted pain score. It is not a pain score adjusted for activity and all other determinants of knee pain. The last concern was related to our assertion that PAKS scores may explain a substantial proportion of low WOMAC scores among patients who have undergone TKR surgery. The possibility that adult patients with knee OA reduce their activity as a means to control pain may explain the observed paradox of low preoperative WOMAC pain scores in patients who undergo TKR. Consistent with this possibility, our study supports the ability of PAKS scores to discriminate the severity of radiographic OA, including situations in which WOMAC scores indicate little or no pain. However, we acknowledge that the PAKS scores need further study before they are appropriate for use as clinical decision-making tools. Knee symptoms are important outcomes in epidemiologic studies and clinical trials. However, focusing only on the WOMAC pain score is inadequate given the paradox of low WOMAC pain scores reported for many adults with radiographic knee OA. The message of our study was to consider an improved method to evaluate symptoms through the combined assessment of knee pain and activity using the PAKS scores.