After stroke, upper extremity (UE) motor impairment is especially disabling and has few effective therapies. Vagus nerve stimulation paired with rehabilitation training (paired VNS) has recently emerged as a promising therapeutic approach. Paired VNS delivers a burst of electrical stimulation to the cervical vagus nerve during a purposeful training movement. In rodent stroke models, paired VNS drives the release of pro-plasticity neurotransmitters that enhance cortical reorganization and boost motor gains. A recent Phase III clinical trial in chronic stroke patients also found that paired VNS improves UE motor impairment compared to rehabilitation without stimulation, but gains were generally modest and clinical benefit varied. It is likely that additional motor gains are possible if paired VNS therapy can be optimized. Understanding how paired VNS works in humans and in whom paired VNS works best will be a critical step in unlocking its clinical potential, allowing us to improve treatment regimens, individualize delivery, and select suitable therapeutic candidates. Preclinical and clinical evidence illuminates several candidate neural substrates for the actions of VNS, but these are yet untested in humans with stroke. Our central hypothesis is that VNS-induced changes in motor, cognitive, and affective systems underlie motor gains, and that sufficient neural integrity is required to achieve a clinically meaningful response. To test this hypothesis, we will study chronic stroke patients implanted with our next- generation VNS device. We will deliver blocks of active or sham VNS paired with UE rehabilitation in a randomized, blinded, crossover design. Before and after treatment blocks, we will characterize neural networks and their associated behaviors using transcranial magnetic stimulation, diffusion and functional MRI, and quantitative behavioral testing. In Aim 1, we will identify VNS-induced neural changes in motor, cognitive, and affective systems that mediate motor gains. In Aim 2, we will identify baseline structural, functional, and injury markers that predict clinical benefit from paired VNS. This project capitalizes on the combined strengths of its multi-PI leadership and investigative team, who bring complementary expertise in motor recovery mechanisms, VNS device development, clinical trial design and execution, and advanced neuroimaging and biostatistical analyses. At the completion of the study, we expect to have generated key mechanistic insights regarding paired VNS in human chronic stroke, providing a solid basis to optimize the efficacy of this therapy. More broadly, this study could reveal targetable pathways for other neuromodulation strategies and could launch new treatments for cognitive and mood disorders after stroke. The proposed research is thus an exciting next step in neurostimulation after stroke. These advances will enhance the clinical potential of paired VNS and will ultimately optimize neurorehabilitation for mill...