<p>This study examined the inhibitory control of response execution when a frequently cued response was occasionally replaced by an alternative one. Healthy participants repeatedly produced a unimanual or bimanual motor response to a frequently presented dummy cue (dummy trials). Occasionally, a signal cue, instructing an alternative response set, was presented instead of the dummy cue (signal trials). The probability of the non-cued response contralateral to the cued response in the dummy trials was lower than that in the signal trials when a unimanual response was cued. This suggests that proactive inhibition of an infrequently cued response is stronger than that of a frequently cued response. The reaction time of the response instructed by the dummy cue was longer when the number of the potentially cued response sets was greater. This indicates that proactive slowing of the frequently cued response becomes more pronounced when the number of the potential response sets is large, owing to the greater information processing demands for proactive inhibition. The reaction time increased when participants occasionally withheld the one-hand response from a frequently cued bimanual response. This indicates that the stopping-interference effect arising from non-selective inhibition occurs when an infrequent cue instructs the withholding of one hand from a frequently cued bimanual response. The time cost of response switching and response addition was not significantly different. In addition, the reaction time in both the frequently cued and infrequently cued responses increased when an infrequent cue instructed an addition of a response to a frequently cued contralateral response. These findings indicate that the stopping-interference effect due to non-selective inhibition occurs even when an infrequent cue does not instruct withholding of a specific response. That is, pausing of the response set in response to an infrequent cue, attributed to non-selective inhibition, occurs regardless of whether it is followed by withholding of a specific response or not. In summary, the present findings highlight that both proactive and non-selective inhibitory mechanisms are flexibly engaged, even in contexts involving infrequent response switching, to ensure robust control of action execution.</p>