A quantile mapping method called RQUANT was applied to the bias correction on the temperature (including daily mean temperature, daily maximum temperature, and daily minimum temperature) simulation of a regional climate model (RegCM4). Assessment on the correction effect was carried out, focusing on the simulation of climatological mean, interannual variation, extreme temperature, and agro-climatic conditions. Results show that this correction method is capable of improving the simulation of mean state of temperature. The biases of the correction results on mean temperature are within 0.5℃. It can also significantly improve the simulation of both extreme temperature and agro-climatic conditions. However, it has little correction effect on interannual variation. Considering the good performance of this method on correcting precipitation simulation in previous study, it implies that the RQUANT method can be used on climate change projection simulations, and these bias corrected simulations would be more useful and reliable for climate change and climate change impact studies.