Abstract: Due to the requirement for prediction of nonlinear target trajectory in image sequences, a sliding window kernel ridge regression(KRR) target trajectory predicting algorithm was proposed. The full KRR which posses the bias item was deduced first, and then the iterative form of sliding window KRR algorithm was also derived in this paper. The algorithm was carried out in a sliding way, and the trajectory information in latest frames was used to predict the position in the next frame, which was achieved by using the KRR. The experimental results demonstrate that the proposed algorithm can predict the nonlinear trajectories accurately, and the prediction error is small. The structure of the proposed algorithm is simple and practicable in engineering application.