Aiming at the problem of gliding near space hyper-sonic vehicle (NSHV) trajectory prediction, a trajectory predic-tion method based on aerodynamic acceleration empirical mode decomposition (EMD) is proposed. The method analyzes the motion characteristics of the skipping gliding NSHV and verifies that the aerodynamic acceleration of the target has a relatively stable rule. On this basis, EMD is used to extract the trend of aerodynamic acceleration into multiple sub-items, and aggre-gate sub-items with similar attributes. Then, a prior basis func-tion is set according to the aerodynamic acceleration stability rule, and the aggregated data are fitted by the basis function to predict its future state. After that, the prediction data of the aero-dynamic acceleration are used to drive the system to predict the target trajectory. Finally, experiments verify the effectiveness of the method. In addition, the distribution of prediction errors in space is discussed, and the reasons are analyzed.