Walking with a shorter stride length (SL) was recently proposed for gait retraining in medial knee osteoarthritis; however it was never assessed in this patient population. This study tested the hypothesis that shortening SL while maintaining walking speed reduces knee adduction (KAM) and flexion (KFM) moments in patients with medial knee osteoarthritis. Walking trials with normal SL and SL reduced by 0.10 m and 0.15 m were recorded for 15 patients (10 men, 55.5 +/- 8.7 years old, 24.6 +/- 3.0 kg/m2). SL was modified using an augmented reality system displaying target footprints on the floor. Repeated one-way ANOVAs and post-hoc paired t-tests were performed to compare gait measures between normal and reduced SL. The individual effects of SL reduction were analyzed using descriptive statistics. Group analysis indicated significant decreases in KAM impulse with both SL reductions (p < 0.05). No systematic change was observed in the first peaks KAM and KFM when walking with reduced SL (p > 0.05). Individually, 33 % of the patients decreased the peak KAM, whereas 20 % decreased the KAM impulse. Among these patients with a decrease in peak KAM or in KAM impulse, 0 % and 33 % had a simultaneous increase in peak KFM, respectively. In conclusion, this study showed that SL shortening can decrease kinetic measures associated with the progression of medial knee osteoarthritis in some patients, demonstrating the importance of considering SL modifications on an individual basis. While further research is necessary, notably regarding dose-response relationships and long-term effects, these findings are particularly encouraging because SL reductions could be easily integrated into rehabilitation protocols.