In order to reconstruct trajectory of missile flight test using the test data obtained from telemetry system and trajectory measurement system, a trajectory reconstruction method based on uncented kalman filter (UKF) was presented in this paper. The system state equations and the observation equations were built using the telemetry data and the trajectory measurement data, respectively. The optimal estimation for the trajectory state variables implemented with UKF algorithm. Finally, error characteristic between UKF and traditional extended kalman filter (EKF) the of the reconstruction errors to 3 error sources were analyzed using numerical simulation. The results indicate that UKF shows higher performances in convergence rate, stability and accuracy, and the position errors and the velocity errors are more sensitive to the errors of trajectory measurement, whereas the attitude errors are more sensitive to the initial attitude errors.