How does human brain stimulation result in lasting changes in cortical excitability? Uncertainty on this question hinders the development of personalized brain stimulation therapies. To characterize how cortical excitability is altered by stimulation, we applied repetitive direct electrical stimulation in eight human subjects (male and female) undergoing intracranial monitoring. We evaluated single-pulse corticocortical-evoked potentials (CCEPs) before and after repetitive stimulation across prefrontal ( = 4), temporal ( = 1), and motor ( = 3) cortices. We asked whether a single session of repetitive stimulation was sufficient to induce excitability changes across distributed cortical sites. We found a subset of regions at which 10 Hz prefrontal repetitive stimulation resulted in both potentiation and suppression of excitability that persisted for at least 10 min. We then asked whether these dynamics could be modeled by the prestimulation connectivity profile of each subject. We found that cortical regions (1) anatomically close to the stimulated site and (2) exhibiting high-amplitude CCEPs underwent changes in excitability following repetitive stimulation. We demonstrate high accuracy (72-95%) and discriminability (81-99%) in predicting regions exhibiting changes using individual subjects' prestimulation connectivity profile, and show that adding prestimulation connectivity features significantly improved model performance. The same features predicted regions of modulation following motor and temporal cortices stimulation in an independent dataset. Together, baseline connectivity profile can be used to predict regions susceptible to brain changes and provides a basis for personalizing brain stimulation. Brain stimulation is increasingly used to treat neuropsychiatric disorders by inducing excitability changes at specific brain regions. However, our understanding of how, when, and where these changes are induced is critically lacking. We inferred plasticity in the human brain after applying electrical stimulation to the brain's surface and measuring changes in excitability. We observed excitability changes in regions anatomically and functionally closer to the stimulation site. Those in responsive regions were accurately predicted using a classifier trained on baseline brain network characteristics. Finally, we showed that the excitability changes can potentially be monitored in real-time. These results begin to fill basic gaps in our understanding of stimulation-induced brain dynamics in humans and offer pathways to optimize stimulation protocols.