Correspondence to

i32–i38. Objectives This study compared the thermal, physiological and perceptual responses associated with match-play tennis in HOT (∼34°C wet-bulb-globe temperature (WBGT)) and COOL (∼19°C WBGT) conditions, along with the accompanying alterations in match characteristics. Methods 12 male tennis players undertook two matches for an effective playing time (ie, ball in play) of 20 min, corresponding to ∼119 and ∼102 min of play in HOT and COOL conditions, respectively. Rectal and skin temperatures, heart rate, subjective ratings of thermal comfort, thermal sensation and perceived exertion were recorded, along with match characteristics. Results End-match rectal temperature increased to a greater extent in the HOT (∼39.4°C) compared with the COOL (∼38.7°C) condition (p<0.05). Thigh skin temperature was higher throughout the HOT match (p<0.001). Heart rate, thermal comfort, thermal sensation and perceived exertion were also higher during the HOT match (p<0.001). Total playing time was longer in the HOT compared with the COOL match (p<0.05). Point duration (∼7.1 s) was similar between conditions, while the time between points was ∼10 s longer in the HOT relative to the COOL match (p<0.05). This led to a ∼3.4 % lower effective playing percentage in the heat (p<0.05). Although several thermal, physiological and perceptual variables were individually correlated to the adjustments in time between points and effective playing percentage, thermal sensation was the only predictor variable associated with both adjustments (p<0.005). Conclusions These adjustments in match-play tennis characteristics under severe heat stress appear to represent a behavioural strategy adopted to minimise or offset the sensation of environmental conditions being rated as difficult