The locus coeruleus-norepinephrine (LC-NE) system is a well-established regulator of behavior, yet its precise role remains unclear. Animal studies predominantly support a ''gai'' hypothesis, suggesting that the LC-NE system enhances sensory processing, while human studies have proposed an alternative ''urgenc'' hypothesis, postulating that LC-NE primarily accelerates responses. To address this discrepancy, we administered transcutaneous vagus nerve stimulation (tVNS) in two experiments involving 43 participants. In the first experiment, we showed that 4-second tVNS trains reliably induced greater pupil dilation compared to SHAM condition, indicating increased LC-NE activity. In the second experiment, we applied tVNS during a random dot motion task to assess its impact on perceptual decision-making. Notably, tVNS improved accuracy without affecting reaction times, which appears inconsistent with the ''urgenc'' hypothesis. Drift-diffusion model analyses further supported the ''gai'' hypothesis, revealing that tVNS increased the drift rate, indicative of enhanced evidence accumulation. Accuracy and drift-rate improvements were especially pronounced following errors and in less proficient participants, who otherwise exhibited post-error declines in these measures under SHAM condition. Our findings suggest that the influence of the LC-NE system adapts to task demands, becoming especially beneficial in challenging contexts. Overall, this study underscores the potential of tVNS as a non-invasive tool to investigate the causal role of the LC-NE system in human behavior.