What is the central question of this study? Controlled-hyperthermia heat-acclimation protocols induce an array of thermoregulatory and cardiovascular adaptations that facilitate exercise in hot conditions. We investigated whether this ergogenic potential can be transferred to thermoneutral normoxic or hypoxic exercise conditions. What is the main finding and its importance? We showed that heat acclimation did not affect maximal cardiac output or maximal aerobic power in thermoneutral normoxic or hypoxic conditions. Heat acclimation augmented the sweating response in thermoneutral normoxic conditions. The cross-adaptation theory, according to which heat acclimation could facilitate hypoxic exercise capacity, is not supported by our data. ⋯ Heat acclimation (HA) mitigates heat-induced decrements in maximal aerobic power ( V̇O2peak ) and augments exercise thermoregulatory responses in the heat. Whether this beneficial effect of HA is observed in hypoxic or thermoneutral conditions remains unresolved. We explored the effects of HA on cardiorespiratory and thermoregulatory responses to exercise in normoxic, hypoxic and hot conditions. Twelve men [ V̇O2peak 54.7(standard deviation 5.7) ml kg-1 min-1 ] participated in a HA protocol consisting of 10 daily 90-min controlled-hyperthermia (target rectal temperature, Tre = 38.5°C) exercise sessions. Before and after HA, we determined V̇O2peak in thermoneutral normoxic (NOR), thermoneutral hypoxic (fractional inspired O2 = 13.5%; HYP) and hot (35°C, 50% relative humidity; HE) conditions in a randomized and counterbalanced order. Preceding each maximal cycling test, a 30-min steady-state exercise bout at 40% of the NOR peak power output was used to evaluate thermoregulatory responses. Heat acclimation induced the expected adaptations in HE: reduced Tre and submaximal heart rate, enhanced sweating response and expanded plasma volume. However, HA did not affect V̇O2peak or maximal cardiac output (P = 0.61). The peak power output was increased post-HA in NOR (P < 0.001) and HE (P < 0.001) by 41 ± 21 and 26 ± 22 W, respectively, but not in HYP (P = 0.14). Gross mechanical efficiency was higher (P = 0.004), whereas resting Tre and sweating thresholds were lower (P < 0.01) post-HA across environments. Nevertheless, the gain of the sweating response decreased (P = 0.05) in HYP. In conclusion, our data do not support a beneficial cross-over effect of HA on V̇O2peak in normoxic or hypoxic conditions.