This study investigated the effects of precooling via crushed ice ingestion on cognitive performance during repeated-sprint cycling in the heat. Nine males, non-heat acclimatised to heat (mean age: 28.2 ± 2.7 y; height: 175.7 ± 9.7 cm; body-mass: 76.9 ± 10.6 kg) completed a 30 min bout of repeated-sprint (36 × 4 s sprints, interspersed with 56 s rest-breaks) on a cycle ergometer in a climate chamber (35°C, 70% relative humidity). Crushed ice ingestion (7g·kg^-1, -0.4°C, ICE) or no cooling (CON) interventions were completed at rest, in the climate chamber, 30 min prior to exercise. Working memory was assessed via the serial seven test (S7) and the automated operation span task (OSPAN) at various time points before, during, and post-exercise. Core body temperature (T_c), forehead temperature (T_h), and thermal sensation (TS) were assessed throughout the protocol. Working memory significantly declined during exercise in CON as measured by S7 (p = 0.01) and OSPAN (p = 0.03); however, it was preserved in ICE with no change at the end of exercise in either S7 or OSPAN scores compared to baseline (p = 0.50, p = 0.09, respectively). Following precooling, T_h (-0.59°C, p < 0.001) and T_c (-0.67°C, p = 0.005) were significantly decreased in ICE compared to CON. At the end of the exercise, ICE significantly reduced T_c compared to CON (p = 0.03), but no significant differences were recorded for T_h. Further, TS was lower following precooling in ICE (p = 0.008) but not during exercise. In conclusion, ice ingestion significantly reduced T_h and T_c and facilitated maintenance of cognitive performance during repeated-sprint exercise in the heat, which may lead to better decision making.