Background: Knee osteoarthritis (OA) is a leading cause of chronic pain and disability worldwide. Because walking is recommended to manage symptoms, there is a critical need to develop clinically accessible treatment strategies that minimize the aberrant gait mechanics that are associated with the onset and progression of structural and symptomatic knee OA. The central objective of this dissertation research was to provide foundational insights focused on the development of gait modification strategies that reduce knee joint loading and pain.Methods: Adults with clinically defined knee OA were recruited for Aims 1-2. The relation of tibial acceleration to knee pain and impact loading during walking was assessed in a cross-sectional study (Aim 1). A randomized, parallel-group, feasibility trial evaluated knee pain and impact loading following two walking programs, with and without real-time feedback to reduce tibial acceleration (Aim 2). A secondary analysis using data from the Multicenter Osteoarthritis Study investigated the relation of preferred walking cadence to 2-year changes in knee pain and physical function in adults with or at risk for knee OA (Aim 3). The acute effects of increasing preferred walking cadence on knee joint loading, impact loading, and knee pain were determined in a single-visit repeated measures study (Aim 4).Results: Higher peak tibial acceleration during walking was associated with worse knee pain and higher impact loading. However, a gait retraining program aimed at reducing peak tibial acceleration was no more effective than a standard walking program for reducing knee pain and impact loading. Lower preferred walking cadence was associated with an increased risk of incident frequent knee pain and physical function limitations after 2 years. Increasing preferred walking cadence by 2-10% had no immediate effect on proxy measures of knee joint loads or knee pain, but significantly increased measures of impact loading.iv Conclusions: This dissertation research integrated perspectives from clinical biomechanics, motor learning, and epidemiology to enhance our understanding of tibial acceleration and cadence as modifiable gait parameters and how they relate to lower extremity loading, pain, and physical function among individuals with knee OA. Peak tibial acceleration was identified as a proxy measure of impact loading. Low preferred cadence was identified as a risk factor for longitudinal changes in knee pain and physical function. However, the data did not support the use of a gait retraining intervention to reduce tibial acceleration over a standard walking program for reducing impact loading or knee pain, nor did it support the hypothesized reductions in knee joint moments as a result of increasing cadence. Nonetheless, the findings directly inform the clinical management of knee OA, indicating that reducing peak tibial acceleration and increasing preferred cadence are safe, well-tolerated gait modifications that do not exacerbate knee pain. Future research should consider how these ga...