Urban bus lane is an important measure to promote bus priority strategy, it’s expected to relieve urban traffic congestion and promote green travel. Exploring the influencing factors of bus lane efficiency and how to set up and improve the efficiency of bus lane scientifically is an interesting study concerns. Based on the quantification and grading evaluation of the effectiveness of bus lanes, the random forest regression model is selected to establish the efficiency impact model. Combined with the grey correlation analysis method, this paper analyzes three kinds of influencing factors, including road type, traffic operation condition and bus lane type. Three different types of bus lanes of three-ring Expressway, Beijing-Tibet freeway and Beijing-Hong Kong-Macao Expressway are selected to verify the validity and accuracy of the model. The results show that the design form of bus lane has the greatest influence on its efficiency, and the overall accuracy rate of this influence model in the grading evaluation of bus lane efficiency is 89.1%. The efficiency influence model of urban bus lane can provide important method support for the refinement and scientific setting of bus lane. Keywords：Public transit; bus priority; bus lane; efficiency evaluation; random forest; impact model

CICTP