It is generally assumed that intestinal temperature (Tint), as measured with a telemetric pill, agrees relatively well with rectal temperature (Trec) during exercise. However, whether Tint reflects Trec during prolonged, intense and continuous exercise when cold fluids are consumed is unknown. Therefore, we compared Trec and Tint during a half-marathon during which cold water was ingested to prevent bodyweight (BW) losses >2%. Nine endurance athletes (age 30 ± 5 years) underwent a 21.1 km running time-trial (TT) in the heat (~30 °C and 44% RH) while BW losses were maintained to ~1% with continuous cold (4 °C) water provision. Tint and Trec were monitored throughout the TT. Hypohydration level, TT time and fluid intake were 1.2 ± 0.4% BW, 93.2 ± 9.9 min and 2143 ± 264 ml, respectively. Trec was systematically higher than Tint by 0.25 °C (95% CI: 0.14-0.37 °C). Tint and Trec showed an excellent relative (r = 0.90, p < 0.01), but poor absolute agreement as reflected by a 95% limit of agreement of ±1.07 °C and a standard error of measurement of ±0.39 °C. In conclusion, Tint does not mirror Trec during prolonged, intense running with cold fluid ingestion and, therefore, these measures should not be used interchangeably under this scenario.