Abstract: RCS of aircraft varies with time or radar wave incident angle without regular pattern during its movement. It is necessary to analyze RCS variation using statistic items or models. The research objects of existing literatures are mostly conventional aircrafts. Research on stealth aircraft is poor. A typical stealth aircraft is selected as the research object in this paper. Its dynamic RCS data under different flying attitudes are evaluated by the combination method of physical optics and physical theory of diffraction. And Swerling type Ⅰ and type Ⅲ distribution, chi-2 distribution and lognormal distribution are applied to analyze the statistic characteristics of the RCS datasets. As the ratio of mean to median of stealth target can be less than 1, the complete form of lognormal distribution model is proposed. Lognormal distribution model can provide a better approximation of the RCS datasets than the other models concerning the model errors and the results of fitting goodness test.