In the existing process of maneuvering target tracking, the target’s actual trajectory can not be foreseen precisely. A commonly attempt is under the assumption that the target’s trajectory is prefixed. The prefixed model is usually not inosculated with the target’s actual trajectory, and the target’s tracking precision can not be guaranteed. In this paper a parameter identification model is presented. This model identifies the target’s trajectory dynamically, adapts to the variation of working cycle, and covers kinds of possible state of the maneuvering target. An adaptive filtering is then employed to analyze the tracking. We call it as a whole adaptive tracking in this paper. Simulations verify that this adaptive tracking algorithm effectively improves the tracking precision.