Considering the problem that Pulse Doppler (PD) radar is not able to track target precisely in the presence of velocity gate pull-off/in jamming and clutter environment, an algorithm for target tracking is proposed which is based on the Double Models (DM) and the Amplitude Information (AI). The algorithm establishes two tracking models: one model is based on the position and amplitude measurements, the other is based on the position, velocity, and amplitude measurements. The two models both use Probabilistic Data Association based on AI (AI-PDA) to reduce the influence of clutter as much as possible, then carry out filtering and estimating using the conventional method. If there is no velocity gate pull off/in jamming, the estimations of two models are related on position and velocity; if there is jamming, the estimations are not related. Accordingly, the chi-square test is carried out and the final estimation result is determined after analysis. The simulation results prove the effectiveness of the algorithm.