From time to time, agents have to abandon an action plan immediately before it is executed. The stop-signal task measures stopping performance using stop signal reaction times (SSRTs). It is possible that such performance is influenced by other action plans prepared for later use while the said action is to be stopped. This study combined a stop-signal task with an ABBA sequence: first, an action A is planned; then, an action B is planned; finally, action B is executed before action A. In some trials, action B was to be selectively stopped, as signalled by a stop signal. The Go trials replicated the classic result that performance of action B is impaired if it shares some, but not all, features with action A (partial overlap costs). Importantly, the same pattern emerged for the SSRTs in the stop trials. The performance of stopping action B was impaired when it partially overlapped with the planned action A. Given these results, it seems that the Stop process is subject to similar selection constraints as the Go process. Stopping the execution of an action plan is less efficient if there is potential confusion with another action plan. These results suggest that the stopping mechanism operates on the specific features of an action and could therefore help to refine the mechanism behind action inhibition.