In order to improve the accuracy and reliability of target tracking, the method of combining bearing-only tracking model with nonlinear predictive filtering algorithm was used to determine the real-time location of target. According to simulation, it is found that by this method, the model error can be estimated in real-time and the real-time position of the target can be predicted precisely with appropriate weighted matrix, which indicates that the bearing-only tracking nonlinear prediction filtering algorithm in target tracking applications not only simple for operation, but also has good accuracy and validity.