The classification of rail transit stations has an important impact on the study of passenger flow pat terns, surrounding land use and development trends of different types of stations. Based on AFC data, a variety of effectiveness indicators were used to determine the number of classifications. By way of principal component analysis, k-means clustering and multiple linear regression, qualitative and quantitative analysis were combined to classify the sites. A total of 58 stations of Suzhou Rail Transit Line 1 and 2 were divided into 4 categories, which would lay the foundation for the follow-up study of the classified sites and the development of rail transit.