Research conclusions: The electric efficiency parameter of metro station lighting system is extracted，the set of parameters’hierarchical framework and relationship are clearly defined． Combination of Delphi method and BP neural network method，the index of evaluation system are twice － confirmed in the parameter set． With the analytical hierarchy process and the qusai － entropy method as an analysis tools，the index which affected electric efficiency in subway station lighting system are empowered weights in an integrated way． The electric efficiency evaluation system of whole subway station lighting system which is reflected the expert opinion tendency about energy － saving，which is a scientific management of energy and a comprehensive analysis of energy saving，which is provided the scientific basis for technological transformation of implementation energy － saving and enhancement of energy efficiency．