The effect of humid heat acclimation on thermoregulatory responses to humid and dry exercise-heat stress was studied in six exercise-trained Thoroughbred horses. Horses were heat acclimated by performing moderate-intensity exercise for 21 days in heat and humidity (HH) [34.2-35.7°C; 84-86% relative humidity (RH); wet bulb globe temperature (WBGT) index ~32°C]. Horses completed exercise tests at 50% of peak O₂ uptake until a pulmonary arterial temperature (T_pa) of 41.5°C was attained in cool dry (CD) (20-21.5°C; 45-50% RH; WBGT ~16°C), hot dry (HD 0) [32-34°C room temperature (RT); 45-55% RH; WBGT ~25°C], and HH conditions (HH 0), and during the second hour of HH on days 3, 7, 14, and 21, and in HD on the 18th day (HD 18) of heat acclimation. The ratios of required evaporative capacity to maximal evaporative capacity of the environment (E_req/E_max) for CD, HD, and HH were ~1.2, 1.6, and 2.5, respectively. Preexercise T_pa and rectal temperature were ~0.5°C lower (P < 0.05) on days 7, 14, and 21 compared with day 0. With exercise in HH, there was no effect of heat acclimation on the rate of rise in T_pa (and therefore exercise duration) nor the rate of heat storage. In contrast, exercise duration was longer, rate of rise in T_pa was significantly slower, and rate of heat storage was decreased on HD 18 compared with HD 0. It was concluded that, during uncompensable heat stress in horses, heat acclimation provided modest heat strain advantages when E_req/E_max was ~1.6, but at higher E_req/E_max no advantages were observed.