The drastic increase in the aging population has increased the prevalence of OA inthe United States. The ability to monitor symptoms of OA within a free-livingenvironment and potentially provide an objective measure of pain experienced by those with OA would significantly improve their quality of life. Bioimpedance technology offers anon-invasive sensing solution that could provide physiological information related to theknee such as joint spacing and swelling. This dissertation aims to determine if localized bioimpedance data from knee site tissues can predict the probability of active OA kneepain in older adults. Several additional research aims are addressed in this work. First, a wearable bioimpedance system was designed and validated using commercially availablecircuits for free-living monitoring. Secondly, a signal processing framework was presentedto identify and remove data artifacts that result from data collection in an unsupervised free-living environment. This resulted in the use of the battery voltage, on-board modelmeasurements, and resistance and phase thresholds to ensure adherence to study protocol,confirm sensing operation, and to identify electrode/cable disconnect events. Next, normative impedance values were presented for bioimpedance data collected over theduration of 7 days for older adults. This included identifying the expected range ofimpedance values collected within this population and assessing the variability ofimpedance measurements collected in a free-living environment. Reducing the longitudinal datasets to a single metric for comparisons between healthy and OA groups did not resultin significant differences. This led to the use of a multi-level model, which identified 128kHz resistance and 40 kHz reactance parameters, as significant predictors for theprobability of active OA knee pain.