To solve the problem that the maneuver trajectory of hypersonic vehicle is missing and difficult to predict in practical applications, a trajectory data imputation (TDI) method that combines the K-nearest neighbor (KNN) algorithm and the agglomerative hierarchical clustering (AHC) algorithm is proposed. Firstly, a three degrees of freedom dynamic model is established to solve the data source problem of maneuver trajectory, and the missing maneuver trajectory is processed for the dataset. Secondly, the clustering forest is constructed to search the “nearest” trajectory of the target trajectory. Then, in order to reduce the time for searching nearest neighbor trajectory and reduce errors caused by different distance measurement methods, the KNN algorithm is used to the search and decision process of the clustering forest. Finally, trajectory imputation and trajectory prediction are performed on the missing trajectory dataset with complex maneuvers, and compared with six imputation methods. The results show that the proposed method can better meet the accuracy and time requirements of trajectory imputation and prediction.