Long-term skills training is known to induce neuroplastic alterations, but it is still debated whether these changes are always modality-specific or can be supramodal components. To address this issue, we compared finger-targeted somatosensory-evoked and auditory-evoked potentials under both Go (response) and Nogo (response inhibition) conditions between 10 baseball players, who require fine hand/digit skills and response inhibition, to 12 matched track and field (T&F) athletes. Electroencephalograms were obtained at nine cortical electrode positions. Go potentials, Nogo potentials, and Go/Nogo reaction time (Go/Nogo RT) were measured during equiprobable somatosensory and auditory Go/Nogo paradigms. Nogo potentials were obtained by subtracting Go trial from Nogo trial responses. Somatosensory Go P100 latency and Go/Nogo RT were significantly shorter in the baseball group than the T&F group, while auditory Go N100 latency and Go/Nogo RT did not differ between groups. Additionally, somatosensory subtracted Nogo N2 latency was significantly shorter in the baseball group than the T&F group. Furthermore, there were significant positive correlations between somatosensory Go/Nogo RT and both Go P100 latency and subtracted Nogo N2 latency, but no significant correlations among auditory responses. We speculate that long-term skills training induce predominantly modality-specific neuroplastic changes that can improve both execution and response inhibition.