This study compared cranial (CC) with passive (CON) cooling during recovery on tolerance to subsequent exercise while wearing firefighting protective ensemble and self-contained breathing apparatus in a hot-humid environment. Eleven males (mean ± SD; age, 30.9 ± 9.2 years; peak oxygen consumption, 49.5 ± 5.1 mL·kg−1·min−1) performed 2 × 20 min treadmill walks (5.6 km·h−1, 4% incline) in 35 °C and 60% relative humidity. During a 20-min recovery (rest), participants sat and removed gloves, helmets, and flash hoods but otherwise remained encapsulated. A close-fitting liquid-perfused hood pumped 13 °C water at ∼500 mL·min−1 through the head and neck (CC) or no cooling hood was worn (CON). During rest, neck temperature was lower in CC compared with CON from 4 min (CC: 35.73 ± 3.28 °C, CON: 37.66 ± 1.35 °C, p = 0.025) until the end (CC: 33.06 ± 4.70 °C, CON: 36.85 ± 1.63 °C, p = 0.014). Rectal temperature rose in both CC (0.11 ± 0.19 °C) and CON (0.26 ± 0.15 °C) during rest, with nonsignificant interaction between conditions (p = 0.076). Perceived thermal stress was lower (p = 0.006) from 5 min of CC (median: 3 (quartile 1: 3, quartile 3: 4)) until the end of rest compared with CON (median: 4 (quartile 1: 4, quartile 3: 4)). However, there were no significant differences (p = 0.906) in tolerance times during the second exercise between CC (16.55 ± 1.14 min) and CON (16.60 ± 1.31 min), nor were there any difference in rectal temperature at the start (CC: 38.30 ± 0.40 °C, CON: 38.40 ± 0.16 °C, p = 0.496) or at the end (CC: 38.82 ± 0.23 °C, CON: 39.07 ± 0.22 °C, p = 0.173). With high ambient heat and encapsulation, cranial and neck cooling during recovery decreases physiological strain and perceived thermal stress, but is ineffective in improving subsequent uncompensable heat stress tolerance.