Abstract: For air combat confrontations involving many-vs.-many unmanned aerial vehicles (UAVs), we consider a situation where each side in the engagement has the same number of UAVs and employ influence diagrams (ID) to establish a continuous decision-making process of a multi-UAVs cooperative air combat. We use Bayesian inference to evaluate the air combat situation in real time and use Hungarian algorithm to achieve the target attack assignment for UAVs dynamically. Therefore, the many-vs.-many cooperative air combat problem is transformed into a series of dynamic one-on-one air combat decision-making processes, where a UAV constantly evaluates the combat situation and maneuvers according to the designed decision-making rules. The simulation experimental results show that the designed air combat model is effective.