Purpose: While there are well documented differences in gait between symptomatic knee osteoarthritis (OA) as compared to healthy older adults the mechanism for how these altered movement patterns occur are not well characterized. Quantification of synergistic muscle activation patterns in knee OA may provide insight to the altered control of movement that may result from joint pain, inactivity, or changes in joint structure, and contribute to the decline in physical function. It has been proposed that movement may be controlled through synergistic activation of muscle groups working as a single functional unit. Quantification of these groupings (aka muscle synergies) can provide insight into neural control of movement by describing not only which muscles are coactivating, but also how they work together to carry out functional movements. In both cerebral palsy and stroke, fewer synergies were needed to describe movements, compared to healthy or unaffected limbs, suggesting a simplified control pattern due to co-contraction. In addition, differences found in the make-up of the muscle synergies throughout movement tasks suggest less coordinated movement with more severe pathologies. The purpose of our study is to quantify differences in muscle synergies between individuals with symptomatic knee OA and healthy older adults. We hypothesize that due to cocontraction, there will be fewer muscle synergies in individuals with knee OA compared to healthy older adults, and that the contribution of these synergies will differ throughout the gait cycle. Methods: After IRB approval, 17 healthy older adults (HA) and 17 symptomatic knee OA were enrolled. Participants walked on a treadmill at a preferred pace and following a 2 minute warm-up EMG was collected at 2000 Hz on 8 leg muscles. EMG signals were then filtered, rectified, and normalized to the gait cycle for 9 total strides for each individual. Mean data for each individual was then concatenated to form one group matrix for OA and one for HA in order to account for individual characteristics within groups in synergy analysis. Muscle synergies were extracted using non-negative matrix factorization (NMF) in both HA and OA. Variance accounted for (VAF = 1-SSE/SSTE) was calculated to determine the number of synergies needed to describe >90% of variance in muscle activation and used to test for differences in complexity of movement between groups. Differences in the contribution of individual muscles to each synergy was determined via visual inspection and then the activation of similar synergies at 0, 30, and 66% of the gait cycle (approx. heel-strike(HS), peak knee flexion(pKF), and toe-off(TO) respectively) were compared using one-way ANOVA's. Results: To account for >90% variability, 4 synergies were needed in the OA group (VAF = 93.4%), while only 3 were needed in the HA group (VAF = 92.4%). In order to easily compare between groups 4 synergies were used to describe both OA and HA (VAF = 96.9%). At HS, activation of synergy C (Fig.1) for OA (0.10 ± 0.21) was significantly higher compared to HA (0.06 ± 0.04, P = 0.001). At pKF, activation of synergy B was significantly higher in OA (0.08 ± 0.04) compared to HA (0.05 ± 0.02, P = 0.010). No difference was found between activation of synergies at TO. Conclusions: In contrast to our hypothesis a greater number of synergies were needed to account for >90% variance in the EMG for the OA group compared to the HA group, suggesting that control patterns in OA are not less complex. The differences in the contributions of the VL, VM and MH to synergy B along with greater activations of synergies B and C at pKF and HS respectively in OA suggests in agreement with the hypothesis that there is greater co-activation of the quadriceps and hamstrings in gait. A strong synergistic relationship between the quadriceps and hamstrings in OA may be related to either pain or joint instability and warrants further analysis to understand the contributing factors.