Abstract: Objective: The construction cost of URT (urban rail transit) projects is high. At the planning stage of rail transit projects in cities where the system is not implemented, it is difficult to determine the feasibility of future rail transit construction. Therefore, it is necessary to establish an urban classification model as a reference for predicting passenger flow indicators.Method: Based on PCA (principal component analysis), ICA (independent component analysis), and Hierarchical Clustering, a 2D urban classification model is developed, comprising a general urban characteristic classification submodel and a classification submodel of URT development characteristics. The former extracts the overall characteristics of cities using socioeconomic development data and operational data from existing cities with rail transit system, classifying cities on an annual basis. The latter classifies cities based solely on URT operational data, considering the development trends of rail transit over a specific period. Load matrices and the contribution rates of principal factors are used to select dependent variables, and independent variable combinations are chosen based on goodness of fit. A multivariate linear regression is employed using the mean values of intragroup indicators to establish the relationship between socioeconomic development indicators and URT indicators.Result & Conclusion: The 2D urban classification model enables the classification and ranking of URT cities using available data, assigning them general characteristics and development features. It establishes the relationship between key URT indicators and socioeconomic development indicators. Based on the multivariate linear regression formula, predicted values of URT indicators for some cities where the system is not implemented can be obtained from both the current standpoint and future development perspective, facilitating the evaluation of construction feasibility.