Brain plasticity after stroke remains poorly understood. Patients may improve spontaneously within the first 3 months and then more slowly in the coming year. The first day, decreased edema and reperfusion of the ischemic penumbra may possibly account for these phenomena, but the improvement during the next weeks suggests plasticity phenomena and cortical reorganization of the brain ischemic areas and of more remote areas. Indeed, the injured ischemic motor cortex has a reduced cortical excitability at the acute phase and a suspension of the topographic representation of affected muscles, whereas the contralateral motor cortex has an increased excitability and an enlarged somatomotor representation; furthermore, contralateral cortex exerts a transcallosal interhemispheric inhibition on the ischemic cortex. This results from the imbalance of the physiological reciprocal interhemispheric inhibition of each hemisphere on the other, contributing to worsening of neurological deficit. Cortical excitability is measurable through transcranial magnetic stimulation (TMS) and prognosis has been established according to the presence of motor evoked potentials (MEP) at the acute phase of stroke, which is predictive of better recovery. Conversely, the lack of response to early stimulation is associated with a poor functional outcome. Non-invasive stimulation techniques such as repetitive TMS (rTMS) or transcranial direct current stimulation (tDCS) have the potential to modulate brain cortical excitability with long lasting effects. In the setting of cerebrovascular disease, around 1000 stroke subjects have been included in placebo-controlled trials so far, most often with an objective of promoting motor recovery of the upper limb. High frequency repetitive stimulation (>3 Hz) rTMS, aiming to increase excitability of the ischemic cortex, or low frequency repetitive stimulation (≤1 Hz), aiming to reduce excitability of the contralateral homonymous cortex, or combined therapies, have shown various effects on the functional disability score and neurological scales of treated patients and on the duration of the treatment. We review here the patients’ characteristics and parameters of stimulation that could predict a good response, as well as safety issues. At last, we review what we have learnt from experimental studies and discuss potential directions to conduct future studies.