This study compared two intensities of active recovery on intermittent sprint exercise performance and the development of heat strain in hot, humid conditions. Eight male game players completed four Cycling Intermittent Sprint Protocols (CISP) consisting of twenty 2-min periods, each including 10-s passive rest, 5-s maximal sprint against a resistance of 7.5% body mass and 105-s active recovery. The CISP was performed in mean (S.D.) temperate conditions with active recovery intensities of 50% V˙O2 peak (TEMP50) and 35% V˙O2 peak (TEMP35) and in hot, humid [35.2 (0.4)°C, 80.4 (2.1)% RH] conditions with the same intensities (HOT50 and HOT35, respectively) in a randomised, counterbalanced order. Heat strain (physiological strain index (PSI)) was calculated from rectal temperature and heart rate. All subjects completed the CISP (20 sprints) in TEMP50 and TEMP35. The mean number of sprints completed for HOT50 and HOT35 was 13 (3) and 17 (2), respectively; both of which were lower than TEMP50 and TEMP35 (P<0.01) and different between hot conditions. Reductions in peak power output (PPO) occurred in the TEMP50 and HOT50 by sprint 8 (P<0.05), but in HOT35 a reduction was delayed until sprint 13 (P<0.05). The rate of PSI increase was faster in HOT50 than TEMP50 and HOT35, but peak PSI was not different. By lowering the recovery intensity, one component of the PSI (heart rate) was reduced and intermittent sprint exercise performance was maintained for longer in the heat.