After a stroke, plasticity occurs in the brain from microscopic to network-level with positive but also negative consequences for functional recovery. Why post-stroke plasticity takes a beneficial or a maladaptive direction is still incompletely understood. Because the biological mechanisms underlying sensorimotor learning parallel those observed during recovery, learning mechanisms could be potential modifiers of post-stroke neuroplasticity and have a discrete mal-/adaptive impact on the recovery of sensorimotor function. However, it is not completely clear whether the processes during procedural learning share the same brain circuits with recovery mechanisms and how they relate to recovery of motor function post-stroke. This project aims to characterize the extent to which recovery and procedural learning rely on the same brain circuit dynamic and to further the understanding of the mechanisms underlying adaptive in contrast to maladaptive plasticity during sensorimotor recovery. We propose the innovative implementation of a state-of-the-art analysis method for task-based functional magnetic resonance imaging (fMRI) data for participants with and without sensorimotor impairment. Specifically, using a multivariate pattern analysis approach will serve (1) to identify similarities and differences of network dynamics within defined cortico-subcortical circuits during procedural learning and (2) to link these individual network dynamics to behavioral change in the learning task and descriptors of recovery post-stroke. The project’s expected results will give novel insight into how new sensorimotor information is integrated in the brain after a stroke compared to the intact brain. Identifying brain network alterations during learning in stroke patients could be of potential diagnostic relevance and serve as a marker to guide individualized therapy and the design of future rehabilitative interventions.