This paper carried out the 1982-2014 Chongqing summer climate experiment was by using the coupled model of IAP AGCM4.1 and WRF3.2, and compare the NCEP/DOE reanalysis data driven WRF3.2 (NWRF) downscaling simulation results, and by using the observation data the evaluation of the two trials for simulation of extreme temperature events in Chongqing. The results showed that: (1) two tests are more reproduction of the climatological characteristics of Chongqing summer surface temperature reasonably, IWRF simulation of large deviations, especially for high altitude areas; their simulation of extreme temperature events there is a deviation, the maximum value of the daily maximum temperature (TXx) anomaly distribution, similar to the two. On warm days index (TX90p) and heat wave duration index (HWDI), the performance of IWRF is better than NWRF; in the West and Southeast, two test has the ability to simulate the warm day index; (2) NWRF can simulate the extreme high temperature refers to the interannual variation of the number, and the poor performance of IWRF; (3) they can simulate the variation trend of TXx, however, NWRF can simulate the increasing trend of HWDI and TX90p. The conclusion of this paper can provide some reference for the short-term climate forecast system of Chongqing for the prediction of extreme high temperature.