Abstract: In order to improve the operating efficiency and promote the sustainable development of public bicycle system, to solve the mismatching problem between the scale of public bicycle and the demand of use, a model to optimize public bicycle station scale was proposed. Based on the IC-card data and station monitoring data, the walking attractive radius of station and the influencing factors of station scale were analyzed by conducting field research and questionnaire survey. The station scale optimization model based on the distribution of generated and attractive points was put forward, following the principles of distance decay law, superimposed effect and balance law. In addition, calibration and test work were carried out with actual data of Beijing. The verification result show that the model precision is 86%. It means that the station scale can be calculated well based on the distribution of generated and attractive points. The results provide important theoretical support for the station planning and optimization of public bicycles.