Intermittent exercise performance in a hot environment is lower than in temperate conditions. Cooling strategies at rest are important; however, the appropriate cooling temperature is not clear. The purpose of this study was to investigate the effect of temperature for cooling leg during half-time (HT) on intermittent exercise performance. Eleven men performed two pre-tests and three experimental trials of a 2 × 30 min intermittent exercise protocol in the heat (33°C; 50% relative humidity). During HT, three experimental conditions were set: no cooling (CON), cooling at the thigh using 12°C ice packs (COOL), and cooling at the thigh using 0°C ice packs (ICE) for 15 min. During the 2nd half, which consisted of 5 s maximal power pedaling every minute separated by 50 s of unloaded pedaling (80 rpm) and rest (5 s), the subjects repeated 2 × 15 sets of the intermittent exercise protocol. The peak power output under the COOL condition (694 ± 80 W) was significantly higher than that of the CON (653 ± 84 W). Thigh skin and deep temperatures showed a significant difference between each condition. In the COOL, thermal sensation was significantly lower, and thermal comfort was significantly higher. In a protocol that simulates intermittent athletic competitions such as soccer in a hot environment, using ice packs at 12°C to cool the thigh during HT lowered thigh skin and deep temperatures and improved heat perception and intermittent exercise performance. This cooling strategy is practical, suggesting that it would be effective for actual competitive situations.