Purpose <p>To investigate whether mental fatigue induced by psychological tasks during whole-body surface cooling influences subsequent endurance exercise performance and whether subjective fatigue is associated with performance variability and neuroendocrine responses.</p> Methods <p>Nine young adults participated in two trials: a mental stress (MS) trial involving the Stroop test (500 stimuli/set × 2) and a control (CON) trial involving watching a video. Both were conducted during whole-body surface cooling using a water-perfused suit (10&#xa0;°C, 10&#xa0;min, then 15&#xa0;°C, 85&#xa0;min). The time to exhaustion (TE) during cycling at 80% peak oxygen uptake was evaluated. Chalder’s fatigue score (CFS), plasma adrenaline ([Ad]_p), noradrenaline ([Norad]_p), and cortisol ([Corti]_p) concentrations were measured before and after intervention and exercise.</p> Results <p>After the intervention, the CFS increased in the MS trial, but not in the CON trial. [Ad]_p and [Norad]_p remained unchanged after the intervention and increased after exercise in both trials, and an increase with exercise was prominent in the MS trial. [Corti]_p remained unchanged. TE decreased by 5.7% in the MS trial. However, this was not statistically significant. The difference (Δ) in TE between trials was significantly negatively correlated with ΔCFS (<i>r</i> = −&#xa0;0.919, <i>P</i> = 0.002). ΔCFS was significantly positively correlated with Δ[Norad]_p (<i>r</i> = 0.795, <i>P</i> = 0.010), while not with [Ad]_p or [Corti]_p.</p> Conclusion <p>Mental fatigue accompanied by whole-body surface cooling variably influences subsequent endurance exercise performance with decrements associated with greater subjective fatigue and heightened sympathoadrenal–medullary activity though trial-level effects were not significant. There was no apparent hypothalamic–pituitary–adrenal axis involvement.</p>