Recent research on the radiographic imaging of knee OA has helped clarify the features of imaging protocols that contribute to accurate representation of disease severity--specifically, the thickness of articular cartilage--and to sensitive detection of disease progression. The absence of standards for reproducible positioning of the knee in the conventional standing AP view obscures the true rate and variability of JSN in knee OA. Moreover, the standing AP view is susceptible to systematic bias insofar as longitudinal changes in knee pain might lead to over- or underestimation of radiographic JSW depending on the direction of change in pain. More recent protocols for standardized knee radiography have been designed to achieve reproducible alignment of the medical tibial plateau and x-ray beam. As a group these protocols permit measurement of tibiofemoral JSW with remarkable precision--the sine qua non of sensitivity to change--however, only limited longitudinal data is available to permit a direct evaluation of the suitability of these protocols for use in clinical DMOAD trials. Longitudinal studies published to date suggest that fluoroscopic positioning methods are superior to nonfluoroscopic methods with respect to reproducing the position of the knee in serial examinations performed several years apart. Fluoroscopic methods also appear to be superior with respect to achieving parallel alignment of the medial tibial plateau and x-ray beam in serial radiographs, a positioning marker strongly associated with sensitive detection of JSN in knee OA. It is important to note that while the various standardization protocols described in this article perform with great success in short-term demonstrations of the reproducibility of positioning and radiographic JSW, differences clearly exist between protocols in the quality of performance over intervals relevant to clinical DMOAD trials. Over intervals of 2 to 3 years, changes in patient characteristics (e.g., severity of knee pain, body weight, load bearing, varus--valgus deformity) and uncontrollable events related to radiography (e.g., technologist turnover, equipment upgrades) have ample opportunity to affect the technical quality of a radiological knee examination. It is difficult, therefore, to conclude whether or not an apparent difference with respect to sensitivity to OA progression between specific radiographic protocols, implemented in separate locations with different cohorts, reflects a robust difference in technical quality or uncontrollable patient variables and events. The most informative recent studies have provided the results of head-to-head longitudinal comparisons of alternative standardization protocols or conventional examination methods performed concurrently in the same subjects [20,22]. Additional comparative studies of this nature are needed, however, to fully characterize the strengths and weaknesses of currently available alternatives in a way that will permit generalizable conclusions regarding the best radiographic methods for multicenter DMOAD trials.