Accurate simulation of the spatial distribution of atmospheric pollutant PM2.5 is the basis of the air pollution control. This study uses PM2.5 concentration of 13 monitoring sites in Xi`an of China based on geographic information system (GIS) and Spearman correlation, application of inverse distance weighted (IDW), ordinary kriging (OK) and trend surface (TS) method, conducted the spatial interpolation analysis of PM2.5 concentration and comparison of the accuracy of different interpolation method. The results show that the accuracy of the IDW interpolation is the highest, the mean error (ME), the mean absolute error (MAE), the mean relative error (MRE), the root mean square error (RMSE) and the system error (SE) are 0.01, 0.05, 0.01, 0.31 and 0.01 respectively. The OK interpolation method is the lowest and the TS interpolation has higher accuracy. Correlation coefficient of simulated and observed value is 0.99 by IDW interpolation, while OK and TS interpolation are 0.62 and 0.67 respectively. The higher PM2.5 concentration areas distributed in the northern, southern and northeast while lower concentration zone located in the western and southeast