Aimed at the problems that the risk of collision between military and civil aviation in the airspace is increasing sharply, the current flight conflict detection methods are difficult to grasp the overall conflict situation in the airspace, and the controllers fail to make accurate judgment of different conflict situations, a method for identifying flight conflict key nodes based on complex network theory is proposed. Firstly, a flight conflict situation network model is built based on ACAS protected area model. Then the importance of all conflict nodes is evaluated based on the centrality of node degree and closeness, PageRank in complex network theory and AHP method to evaluate nodes’ conflict levels. And the key aircraft and locations with high threat levels are found. The simulation results show that the level of security situation in the airspace can be divided reasonably by establishing the flight conflict situation network. Simultaneously, according to the evaluation index of node importance of complex network, aircraft with serious conflict security threats can be effectively identified to assist air traffic controllers to fully grasp the flight safety situation in the airspace.