In order to solve the shortcomings of the existing methods in dealing with uncertainty information reasoning, a dynamic Bayesian network model for threat assessment is proposed. The state transition probability table and conditional probability table of the nodes in the network are given based on analysis of dynamic Bayesian network and the analysis of the influencing factors of air strike target threat. The posterior probability of the target threat is obtained by reasoning with the evidence observations, and the threat level of the typical airway air attack target is simulated and analyzed. The results show that the model is reasonable and effective, and can accurately reflect the real threat level of the targets, thus effective terminal interception is implemented for air attack targets.