Abstract: Based on interference events that are influencing metro system vulnerability, directional complex network model of interference events is constructed. By using analytical indicators including network density, average route length, degree and degree distribution, and clustering coefficient, topological characteristics analysis is conducted on complex network. Analytical results show that only few key nodes can directly affect metro vulnerability. By adopting mutual information theory of complex network, information calculation and importance evaluation are conducted for interference events directional complex network nodes, and to further recognize important nodes. Evaluation results show that the key factors influencing metro system vulnerability is the station passenger flow volume, vehicle system and equipment system, and corresponding measures can be adopted to improve metro operation safety.