The effect of localised head and neck per-cooling on central and peripheral fatigue during high thermal strain was investigated. Fourteen participants cycled for 60 min at 50% peak oxygen uptake on 3 occasions: thermoneutral control (CON; 18 °C), hot (HOT; 35 °C), and HOT with head and neck cooling (HOT cooling ). Maximal voluntary force (MVF) and central activation ratio (CAR) of the knee extensors were measured every 30 s during a sustained maximal voluntary contraction (MVC). Triplet peak force was measured following cycling, before and after the MVC. Rectal temperatures were higher in HOT cooling (39.2 ± 0.6 °C) and HOT (39.3 ± 0.5 °C) than CON (38.1 ± 0.3 °C; P < 0.05). Head and neck thermal sensation was similar in HOT cooling (4.2 ± 1.4) and CON (4.4 ± 0.9; P > 0.05) but lower than HOT (5.9 ± 1.5; P < 0.05). MVF and CAR were lower in HOT than CON throughout the MVC ( P < 0.05). MVF and CAR were also lower in HOT cooling than CON at 5, 60, and 120 s, but similar at 30 and 90 s into the MVC ( P > 0.05). Furthermore, they were greater in HOT cooling than HOT at 30 s, whilst triplet peak force was preserved in HOT after MVC. These results provide evidence that central fatigue following exercise in the heat is partially attenuated with head and neck cooling, which may be at the expense of greater peripheral fatigue. Novelty Central fatigue was greatest during hyperthermia. Head and neck cooling partially attenuated the greater central fatigue in the heat. Per-cooling led to more voluntary force production and more peripheral fatigue.