Distance running in heat and humidity elevates body core temperature (Tc), limits heat dissipation, and increases the risk of heat illness (Armstrong et al. 1996). Post-race rectal temperature has traditionally quantified Tc rise during outdoor running (Cheuvront & Haymes, 2001). The invasive, obtrusive, and impractical nature of rectal thermometry has restricted continuous measurement of Tc during outdoor running to a few case-studies (e.g. Maron et al. 1977). The present study aimed to continuously measure Tc using ingestible telemetric temperature-sensing pill technology in a sample of runners competing in a 21-km race in environmental conditions representing a high-risk of heat illness. Volunteers were 18 heat acclimatized male soldiers (aged 20-23 years) from the Singapore Armed Forces participating in the 2003 Singapore Army Half-Marathon. Each volunteer ingested a telemetric temperature-sensing pill (8-10 hours before running) and wore an ambulatory data recorder and heart rate (HR) monitor for the continuous measurement of Tc and HR, respectively. Pre to post-race changes in nude body mass quantified fluid balance. Environmental wet bulb globe temperature was 26.0°C at race start (0600 h) and 28.3°C at the time of the last finishing volunteer (0826 h). All volunteers finished the race asymptomatic in a mean ± SEM (range) time of 118 ± 3 (105-146) min. During the initial 30-min of running Tc rate of rise was 3.1 ± 0.2 (1.3-4.0)°Cnull;hr. Ten volunteers achieved peak Tc during the race whereas eight volunteers achieved peak Tc at race finish. Peak Tc was 40.1 ± 0.2 (39.3-41.7)°C at 86 ± 8 (13-130) min and final Tc was 39.9 ± 0.2 (38.3-41.7)°C. All volunteers demonstrated Tc > 39°C; 56% (n=10) Tc > 40°C; and 11% (n=2) Tc > 41°C. Three general patterns of Tc response demonstrating large inter-subject variability were observed: 1) increase followed by steady state (n=7); 2) continual increase until race finish (n=5); and 3) increase followed by marked decline (n=6). The rate and magnitude of Tc response was unrelated (Pearson correlation coefficient, P > 0.05) to average running velocity for 21-km or any fluid balance variable (e.g. % dehydration, % sweat loss replaced). Peak Tc was significantly and positively related to peak (r = .72, P < 0.01) and mean (r = .57, P < 0.05) HR. Telemetric temperature pill technology demonstrated significant practical utility for continuous Tc measurement in unrestricted ambulatory subjects. Temperature responses demonstrated large inter-subject variability in magnitude and temporal nature, which were unrelated to fluid balance and race time.