The aviation noise caused by the huge passenger and cargo volume of Taoyuan International Airport has a major impact on the quality of the local residents' life. Aviation noise pollution is one of the important external costs of airports. This study applied the data of ''Real House Value Registry Scheme'' in noise control areas from 2016 to 2018 and the data of 18 aviation noise monitoring stations. Since the value of the noise monitoring station is easily affected by location and distance, this study proposes an estimated aviation noise as one of the explanatory variables. The noise prevention fee is also used as an explanatory variable. This study uses four different hedonic pricing models including linear, semi-logarithmic, anti-log and double logarithmic, under the consideration of spatial autocorrelation of data points, to explore the impact of aviation noise on real estate prices and the cost of airport externalities. Empirical results show that, among all the constructed models, the geographically weighted regression model with linear model has the best explanatory power. The model performance of using the proposed new estimated aviation noise as the explanatory variable are all better than that of using the traditional monitoring station noise. The results also revealed that aviation noise has a significant impact on the real estate prices of noise control areas, but noise prevention fee is not significant which means the current subsidy policy is at an appropriate level.