To make the environmental performance measuring system a useful inspective and supervisory tool during the life cycle of a Green Manufacturing (GM) program, a measure system is presented. It includes three primary indexes corresponding to the GM program life cycle, which are fundamental construction level, application level and continuous improving level. And 19 secondary measures are subdivided, too. After that, the values of the measures are determined involving qualitative and quantitative index. And then Back Propagation Neural Network (BPNN) is applied to evaluate the environmental performance of GM programs in enterprise, in which 10 samples are used to train the measuring model with Matlab 7.0, and the data are selected from the Internet and digital library. The environmental performance evaluation of a machine tool company in Chongqing shows the effectiveness and validity of the model, and this method can be applied in industrial enterprises while implementing GM.