In the warships formation air defense operations, in the process of threat assessment, there are many problems, such as many uncertainties, lack of observation data or errors, and the existing assessment methods are difficult to consider the dynamic threat situation. A threat assessment method based on discrete dynamic Bayesian networks cloud (DDBN-cloud) model is proposed. By analyzing the characteristic of incoming targets, an assessment system is constructed. The continuous variables in the network structure are transformed into discrete variables by using fuzzy logic theory, which can avoid the repeated computation caused by the continuous change of threat attribute values of nodes in a small range. Aiming at the problem of missing data in the evaluation process, a forward information repair algorithm is proposed to carry out information prediction and repair. The prior probabilities of uncertain nodes are assigned according to the reliability of evidence, which makes the parameters of BN more realistic. Finally, the Cloud model is used to transform the probability of the threat assessment to the value of the threat assessment, and then threat ranking is carried out. Simulation results show that the proposed method is suitable for dynamic threat assessment when the target data is missing. Compared with the static Bayesian networks cloud (BN-cloud) model method and the relative entropy ranking method, the proposed method is more reasonable and has certain practical value.