Context: Knee osteoarthritis (OA) is a common and debilitating source of pain and disability in the United States. Primary care providers are the first line in diagnosis and treatment, which walking is a mainstay of therapy. Many knee OA patients do not exercise enough, but their neighborhood could present both a barrier and a solution. Objective: To investigate whether knee OA severity is clustered in neighborhoods and which factors may drive that distribution. Study Design: Cross-sectional. Setting or Dataset: Primary care patients at an academic family and internal medicine practice. Population studied: Established adult patients aged 45-90 with knee OA and no history of total joint arthroplasty in the St. Louis metropolitan area recruited by mail via electronic health record (EHR) data pull. Exposure: Neighborhood determined by geocoded address accounting for built and social environment. Outcome Measures: Geospatial distribution of knee OA cases by severity of OA as measured by the Western Ontario and McMaster University’s OA Severity Scale (WOMAC) via geospatial regression models. Results: Initial recruitment has resulted in a sample of n=77 with mean age 62.5 years and 61% white. Preliminary geospatial analysis revealed clustering of knee OA spatially by severity of OA symptoms (WOMAC). Recruitment continues with another 64 participants added and is ongoing. Conclusions: A multiscale geospatial analysis will investigate further spatial relationships with knee OA severity by WOMAC score accounting for built and social environment. Further analysis anticipated Summer 2021. These finds may help inform primary care providers’ approach to counseling patients with OA on how to leverage their neighborhood to meet their exercise goals and control their knee OA symptoms.