This paper clusters and grades the 53 major cities with high-speed railway based on different evaluation indexes under the existing high-speed railway network in China. Firstly, the correlation analysis of city evaluation indicators was carried out to simplify the process. Four representative categories of evaluation indicators were chosen as the basis for systematic clustering in the end. Secondly, it was found that the distribution of some of the cities in certain classes was too centralized after the city-sites were classified into three grades by using systematic clustering with SPSS. Finally we introduced geographical factors by the k-means clustering to reclassify the cities and obtained the final classification result. This study provides the basis and theoretical support for the sites selection problem of high-speed railway express transit sites in the context of hub-and-spoke network.