The aim of this study was to test the effect of face cooling with cold water (1.2 ± 0.7 ° C) vs. face cooling with neutral water (28.0 ± 3.0 ° C) during high-intensity swimming training on both the core temperature (T co ) and thermal perceptions in internationally ranked long-distance swimmers (5 men’s and 3 women’s) during 2 randomized swimming sessions. After a standardized warm-up of 1,200 m, the athletes performed a standardized training session that consisted of 2,000 m (5 × 400 m; start every 5’15”) at a best velocity then 600 m of aerobic work. Heart rate (HR) was continuously monitored during 5 × 400 m, whereas T co , thermal comfort (TC), and thermal sensation (TS) were measured before and after each 400 m. Before and after each 400 m, the swimmers were asked to flow 200 mL of cold water (1.2 ° C) or neutral (22 ° C) water packaged in standardized bottles on their face. The swimmers were asked don’t drink during exercise. The velocity was significantly different between cold water and neutral water ( p < 0.004 – 71.58 m.min –1 ± 2.32 and 70.52 m.min –1 ± 1.73, respectively). The T co was increased by ±0.5 ° C at race pace, under both face cooling conditions with no significant difference. No significant changes were noted in mean HR (i.e., 115 ± 9 and 114 ± 15 bpm for NW and CW, respectively). TC was higher with Cold Cooling than Neutral Cooling and TS was lower with Cold cooling compared with Neutral cooling. The changes in perceptual parameters caused by face cooling with cold water reflect the psychological impact on the physical parameters. The mean velocity was less important with face cooling whereas the heat rate and T co were the same in the both conditions. The mechanism leading to these results seems to involve brain integration of signals from physiological and psychological sources.