Résumé : Objective: We investigated whether the mid-term impact (1 week posttraining) of a “combined cognitive rehabilitation (CRP)/transcranial direct current stimulation (tDCS) program” on the performance of a Go/No-go task was enhanced compared with isolated CRP and whether it varied according to the stimulation site (right inferior frontal gyrus [rIFG] vs right dorsolateral prefrontal cortex [rDLPFC]). Methods: A total of 150 healthy participants were assigned to (1) an Inhibition Training (IT) group, (2) a group receiving active tDCS over the rIFG in combination with IT (IT + IF), (3) a group receiving active tDCS over the rDLPFC in combination with IT (IT + DL), (4) a group receiving IT with sham tDCS (ITsham), and (5) a No-Training (NT) group to control for test-retest effects. Each group undertook 3 sessions of a Go/No-go task concomitant with the recording of event-related potentials (T0, before training; T1, at the end of a 4-day training session [20 minutes each day]; T2, 1 week after T1). Results: With the exception of the NT participants, all the groups exhibited improved performances at T2. The IT + DL group exhibited the best improvement profile, indexed by faster response times (RTs) (T0 > T1 = T2), with a reduced rate of errors at the posttraining sessions compared with both T0 and T1. This “inhibitory learning effect” was neurophysiologically indexed by shorter No-go N2d latencies and enhanced No-go P3d amplitudes. Conclusion: CRP combined with active tDCS over the rDLPFC appears to be optimal for boosting long-term (one week) inhibitory skills as it induced specific and robust neural changes.