Increasing organizational dependence on the IS has led to a corresponding increase in the impact of information systems security (ISS) abuses. In this paper, we propose a multi-target synthetic assessment approach for ISS risk based on neural network. Firstly, an assessment system of ISS risk is established. Secondly, a trained neural network can endue the assessment approach with expert assessment knowledge in the way of joint weight, through its learning and adapting capability. Finally, the approach is capable to simulate expert assessment method by verifying the data of ISS risk assessment.