We used incorrect visual feedback of ambient and core temperature in the heat to test the hypothesis that deception would alleviate the decrement in cycling performance compared to a no deception trial. Seven males completed three 30 min cycling time trials in a randomised order on a Kingcycle ergometer. One time trial was in temperate, control conditions (CON: 21.8 ± 0.6°C; 43.3 ± 4.3%rh), the others in hot, humid conditions (HOT: 31.4 ± 0.3°C; 63.9 ± 4.5%rh). In one of the hot, humid conditions (31.6 ± 0.5°C; 65.4 ± 4.3%rh), participants were deceived (DEC) into thinking the ambient conditions were 26.0°C; 60.0%rh and their core temperature was 0.3°C lower than it really was. Compared to CON (16.63 ± 2.43 km) distance covered was lower in HOT (15.88 ± 2.75 km; P < 0.05), but DEC ameliorated this (16.74 ± 2.87 km; P < 0.05). Mean power output was greater in DEC (184.4 ± 60.4 W) than HOT (168.1 ± 54.1 W; P < 0.05) and no difference was observed between CON and DEC. Rectal temperature and iEMG of the vastus lateralis were not different, but RPE in the third minute was lower in DEC than HOT (P < 0.05). Deception improved performance in the heat by creating a lower RPE, evidence of a subtle mismatch between the subconscious expectation and conscious perception of the task demands.