The contribution of movement-related activity to Go/NoGo ERP differences has been debated for 25 years. In this study, we examined ERP and fMRI measures of activity in twenty adults performing non-motor (count) and motor (right-handed button press) trials of the Go/NoGo task. Task performance was highly accurate and similar in the ERP and fMRI environments. No significant task-related effects were observed for the N2 component; however, we observed a substantial increase in positivity for Press NoGo compared to Count NoGo trials. The fMRI results also revealed significant deactivations for Press NoGo relative to Count NoGo trials in several left-lateralised motor-related areas, including the inferior frontal gyrus, precentral gyrus and supplementary motor area. Together, the results indicate that the P3 NoGo>Go effect in motor tasks is caused not by movement-related negativity on Go trials but by inhibition-related positivity on NoGo trials, and that this is associated with deactivation of motor areas involved in the Go response.