Generalization in motor control is defined as the ability to carry over information from trained experiences to novel situations. The generalization capacity is critical for the efficacy of robotic-assisted rehabilitation. Namely, if devices like exoskeletons or treadmills are to be used as training devices, patients must generalize the movements learned with the devices to real-life situations without them. Therefore, there is an interest in finding factors facilitating the generalization of corrected movements on a training device beyond the clinical setting. My dissertation focuses on identifying factors regulating the generalization process in locomotor adaptation. To this end, I used a split-belt treadmill (training context) to induce locomotor adaption by moving the legs at different speeds, and I evaluated the generalization of the treadmill-adaptation effects (after-effects) to overground walking (testing context). I specifically determined the extent to which small vs. large perturbations during split-belt walking have a distinct effect on the generalization of adapted movements in young and older adults (Aim 1). I found that older adults generalize their movements more than young regardless of the perturbation size experienced during adaptation. I also investigated the impact of increasing the extent of treadmill-adaptation vs. reducing the contextual similarity between the treadmill and overground walking on the generalization of locomotor adaptation (Aim 2). Results from this aim showed that contextual similarity is more important than the extent of adaptation in the generalization of corrected movements. Thus, in my last aim I used a pair of motorized shoes to induce split-belt treadmill-like adaptation that could increase the contextual similarity between walking with the motorized shoes (training context) and walking without them (testing context) (Aim 3). Results from Aim 3 confirmed that a pair of motorized shoes can induce the same type of robust locomotor adaptation as the split-belt treadmill, opening the possibility to enhance the generalization of correct movements with this training device to walking without it. Taken together, my work advanced our understanding of generalization in locomotor adaptation and has the potential to guide training strategies exploiting the human generalization ability to benefit motor performances in new situations.