This study aims to investigate the extreme temperature variations of asphalt pavement structures. First, this study established a thermal analysis model using finite element methods to investigate the temperature profile of asphalt pavements. Thereafter, the significant factors affecting the pavement temperature were determined using the grey correlation method. Finally, linear prediction models for maximum, minimum, and average temperatures of asphalt layers at any depth were developed. The results indicated that the pavement and air temperature change trends were similar, and that the temperature decreased with increasing depth. The maximum temperature occurred on the road surface during the daytime and moved from the asphalt layer to the base layer during the night. The maximum temperature difference on the road surface was 26.8 °C. The most significant factors affecting the asphalt pavement temperature were air temperature and solar radiation. The maximum, minimum, and average temperatures of the asphalt layers had good linear relationships with air temperature, solar radiation, and depth of concern, and the R 2 of these predicted models were all greater than 0.94. In addition, a comparison between the predicted and actual measured pavement temperature of Beijing’s RIOHTRACK road in 2019 found that the R 2 between the measured and predicted values was greater than 0.90 and the root-mean-squared error was less than 5 °C. Using this model, the maximum, minimum, and average temperatures of the asphalt layer at any depth can be determined if the air temperature is known. These results will be beneficial for analyzing the stresses and strains resulting from temperature differences and calibrating the structural design methods for asphalt pavements.