Using the monitoring data from the automatic weather stations of Sanchakou bridge(altitude of 708 meters)near Enshi on Huyu freeway from June to August 2013(summer) and from December 2013 to February 2014(winter),the variation of road surface and air temperatures on Sanchakou bridge were analyzed statistically, and the relationship between the maximum temperature in summer and the minimum temperature in winter of road surface and meteorological factors were studied. Then, prediction equations of the road surface temperature were established by using multiple regression analysis, and the accuracies of prediction were tested. Results showed that the maximum road surface temperature in summer was influenced by a variety of meteorological factors. The maximum road surface temperature had a positive correlation with the maximum and the mean road surface temperatures on the day before, and the maximum and the mean air temperature, while they had a negative correlation with relative humidity and rainfalls in 24 hours, and the correlation was the most significant with the maximum air temperature. The minimum road surface temperature in winter was significantly correlated with the minimum air temperature. Through the model experiment, the predicted maximum road surface temperature in summer is close to observations while using the forecast model with air temperature, relative humidity and the physical parameters of the previous day. The predicted minimum road surface temperature in winter is close to observations while using the forecast model with air temperature, relative humidity, wind speed and the physical parameters of the previous day.