Reactively canceling movements is a vital feature of the motor system to ensure safety. This behavior can be studied in the laboratory using the stop-signal task. There remains ambiguity about whether a “point-of-no-return” exists, after which a response cannot be aborted. A separate question concerns whether motor system inhibition associated with attempted stopping persists when stopping is unsuccessful. We address these two questions using electromyography (EMG) in two stop-signal task experiments. Experiment 1 ( n = 24) involved simple right and left index finger responses in separate task blocks. Experiment 2 ( n = 28) involved a response choice between the right index and pinky fingers. To evaluate the approximate point of no return, we measured EMG in responding fingers during the 100 msec preceding the stop signal and observed significantly greater EMG amplitudes during failed than successful stopping in both experiments. Thus, EMG before the stop signal differentiated success, regardless of whether there was a response choice. To address whether motor inhibition persists after failed stopping, we assessed EMG peak-to-offset durations and slopes (i.e., rate of EMG decline) for go, failed stop, and successful stop (partial response) trials. EMG peak-to-offset was shorter and steeper for failed stopping compared to go and successful stop partial response trials, suggesting motor inhibition persists even when failing to stop. These findings indicate EMG is sensitive to a “transition zone” at which the relative likelihood of stop failure versus success inverts and also suggest peak-to-offset time of response-related EMG activity during failed stopping reflects stopping-related inhibition.