To solve the problems that some objects' attributes contain preference information and noise may exists in the objects' attributes or the attributes are incomplete, based on the analysis of traditional rough sets theory, a dominance variable precision rough sets method is put forward and the concrete steps of the dominance variable precision rough sets algorithm are given. According to the UCAV's combat characteristics, the algorithm is applied in the process of UCAV's threat assessment. A decision-making information system of UCAV's threat assessment based on the dominance variable precision rough sets method is brought forward, which is demonstrated in an example. Condition attributes and decision-making attributes in the decision-making system are also given. The result shows that the algorithm is a good solution to the problems above and the threat grade of the target is assessed properly. Decision-making rules are obtained with certain credibility, which ensures the consistency of the rules. Regarding the decision-making information system with preference attributes, the incompatibility among the rules can be distinguished by using this method.