Response inhibition is key to flexible behavior. Importantly, performance in any task, including response inhibition tasks, fluctuates on a moment-to-moment basis. Using pupillometry, we investigated the relationship between these behavioral fluctuations in response inhibition and naturally occurring fluctuations of norepinephrine (NE) levels in the brain before a given trial has even started. This was motivated by earlier pharmacological work suggesting a pivotal role of NE in response inhibition, in particular. We specifically used two pupillometry proxies for pretrial (tonic) NE levels, the absolute pretrial pupil size and its derivative, and investigated whether and to which degree they were related to response-inhibition performance in a stop-signal task. Specifically, we investigated the relationship to stopping success, and the speed of the go response (GoRT) and that of the stop response (SSRT). In two experiments, we showed that larger pretrial pupil measures predicted (1) lower stopping success, (2) faster GoRTs (particularly so when the go response needed to be executed in a stop context), and some evidence for (3) faster SSRTs. Taken together, our findings show a clear pattern that pretrial pupil measures predict behavioral fluctuations in response inhibition, which suggests that tonic levels of NE are involved in the regulation of these behavioral fluctuations. Yet, our work furthermore indicates that this involvement is not stopping-specific, given its effect on both the go and the stop response.