To improve tracking performance in bearings-only target tracking problems, a maneuver-adaptive filtering algorithm was proposed in this work. The proposed algorithm based on “current” statistical model adjusted maneuver indices (maneuver frequency and extrema in random acceleration distribution) in the model adaptively by monitoring the filter’s innovation sequence. Thus, the maneuver-adaptive filtering algorithm was able to adjust object adaptively in accordance with practical target motion situation and track time-variant maneuvering target effectively. Simulation results show that the proposed algorithm can achieve high tracking accuracy against weak maneuvering target and high tracking ability against strong maneuvering target.