The effects of dry-bulb temperature on self-paced exercise performance, along with thermal, cardiovascular, and perceptual responses, were investigated by minimizing differences in the skin-to-air vapor pressure gradient (Psk,sat - Pa) between four temperatures. Fourteen trained male cyclists performed 30-km time trials in 13°C and 44% relative humidity (RH), 20°C and 70% RH, 28°C and 78% RH, and 36°C and 72% RH. Power output was similar in 13°C (275 ± 31W; means and SD) and 20°C (272 ± 28W; P = 1.00), lower in 36°C (228 ± 36W) than 13°C, 20°C, and 28°C (262 ± 27W; P < 0.001) and lower in 28°C than at 13°C and 20°C (P < 0.001). Peak rectal temperature was higher in 36°C (39.6 ± 0.4°C) than in all conditions (P < 0.001) and higher in 28°C (39.1 ± 0.4°C) than 13°C (38.7 ± 0.3°C; P < 0.001) and 20°C (38.8±0.3°C; P < 0.01). Mean heart rate was higher in 36°C (163±14 beats min-1) than all conditions (P < 0.001) and higher in 20°C (156 ± 11 beats min-1; P = 0.009) and 28°C (159 ± 11 beats min-1; P < 0.001) than 13°C (153 ± 11 beats min-1). Mean cardiac output was lower in 36°C (16.8 ± 2.5 L min-1) than all conditions (P < 0.001) and lower in 28°C (18.6 ± 1.6 L min-1) than 20°C (19.4 ± 2.0 L min-1; P = 0.004). Ratings of perceived exertion were higher in 36°C than all conditions (P < 0.001) and higher in 28°C than 20°C (P < 0.04). Self-paced exercise performance was maintained in 13°C and 20°C at a matched evaporative potential, impaired in 28°C, and further compromised in 36°C in association with a moderately lower evaporative potential and marked elevations in thermal, cardiovascular, and perceptual strain.