The motor system has the critical capacity to adapt actions through interactions with the world. Two processes that contribute to this capacity are 1) the recalibration of internal representations of the body and environment for predictive motor control and 2) the switching between these internal representations. My dissertation focuses on identifying factors regulating these two processes in locomotion. To this end, I used a split-belt treadmill to create novel walking environments by moving the legs at different speeds, which has been demonstrated to induce sensorimotor recalibration in locomotion. I specifically determined the impact of kinetic demands during split-belt walking on the recalibration of movements (Aim 1), the effect of sensorimotor recalibration on the estimation of limb position across different walking conditions (Aim 2), and the influence of cognitive processes on the generalization of recalibrated movements across different environments in older and younger adults (Aim 3). Results from Aim 1 indicate that the extent of motor adaptation in the novel environment, and subsequent recalibration, is strongly regulated by the specific kinetic demands imposed by the split-belt condition. This finding suggests that manipulation of kinetic demands could augment the adaptation and recalibration of post-stroke gait asymmetry. Results from Aim 2 indicate that the estimation of limb position changes after split-belt walking, regardless of the walking speeds at which it is tested. This indicates that the recalibration of internal representations for walking has an impact on the predictive control of our actions and on subjects’ perception of their movements. These results suggest that sensorimotor recalibration could potentially improve patient’s perception of step length asymmetry, which is often inaccurate after stroke. Lastly, results from Aim 3 indicate that cognitive processes that contribute to action selection also influence the generalization of recalibrated movements across environments in older adults. This is important because it indicates that switching between distinct internal representations of the environment might be an automated process that becomes more consciously controlled with healthy aging. In conclusion, my work advanced our understanding of locomotor adaptation and has the potential to guide gait rehabilitation strategies exploiting sensorimotor adaptation processes.