The affect of hyperthermia on cognitive function remains equivocal, perhaps because of methodological discrepancy. Using electroencephalographic event-related potentials (ERPs), we tested the hypothesis that a passive heat stress impairs cognitive processing. Thirteen volunteers performed repeated auditory oddball paradigms under two thermal conditions, normothermic Time control and Heat stress, on different days. For the Heat stress trial, these paradigms were performed at pre-heat stress (i.e., normothermic) baseline, when esophageal temperature (Tes) had increased by ~0.8°C, when Tes had increased by ~2.0°C, and during cooling following the heat stress. The reaction time and ERPs were recorded in each session. For the Time control trial, subjects performed the auditory oddball paradigms at approximately the same time interval as they did in the Heat stress trial. The peak latency and amplitude of an indicator of auditory processing (N100) were not altered regardless of thermal conditions. An indicator of stimulus classification/evaluation time (latency of P300) and the reaction time were shortened during heat stress, moreover an indicator of cognitive processing (the amplitude of P300) was significantly reduced during severe heat stress (8.3±1.3μV) relative to the baseline (12.2±1.0μV, p<0.01). No changes in these indices occurred during the Time control trial. During subsequent whole body cooling, the amplitude of P300 remained reduced, and the reaction time and latency of P300 remained shortened. These results suggest that excessive elevations in internal temperature reduce cognitive processing but promote classification time.