Regarding the difficulties in the assessment of innovation and entrepreneurship teaching capability currently like biased qualitative assessment and less nonlinear quantitative assessment and insufficient assessment efficiency, an assessment model of innovation and entrepreneurship education targeted at local University students on the foundation of the Back Propagation (BP) neural network is designed. First, the contents of the questionnaire are carefully designed, on whose basis the assessment index system is built. Then, a BP neural network model is utilized for the purpose of comprehensively evaluating the collected data. Lastly, we carry out a demonstration study on innovation and entrepreneurship education, whose results reveal the students' professional knowledge and skills, participation in national competitions, and centralized entrepreneurship training camp play a key role and have a profound effect on innovation and entrepreneurship capability. The evaluation system is able to effectively assess the actual effects of innovation and entrepreneurship education, direct teachers' instructional activity, as well as motivate students to learn and practice.