This study examined whether spontaneous fluctuations in arousal influence the efficiency of exogenous attentional shifts. The Adaptive Gain Theory proposes that arousal affects cognitive performance in a non-linear way, with optimal functioning at moderate arousal levels. Based on this theory, it was hypothesized that attentional orienting would be most effective at intermediate levels of arousal. To test this, forty-six participants completed a spatial cueing task with non-predictive cues while pre-stimulus pupil size was measured as an index of arousal. The difference in response times between validly and invalidly cued targets was used as a measure of attentional orienting efficiency, while overall reaction time served as an indicator of general performance. Results showed a reliable cueing effect: participants responded faster to validly cued targets than to invalid ones. Reaction times also followed a U-shaped pattern across arousal bins, with fastest responses occurring at moderate pupil sizes. This finding is consistent with Adaptive Gain Theory and earlier work on arousal and performance. However, the magnitude of the cueing effect did not vary across arousal levels. These results suggest that while arousal enhances general task performance, it might not influence the efficiency of attentional shifts triggered by external cues. This may indicate that arousal affects only certain components of cognitive performance. The current findings refine our understanding of arousal's role in cognition, indicating that its effects on performance may be selective rather than global, by enhancing general speed without altering the deployment of attention to external stimuli.