The range-Doppler imaging algorithm is a basic imaging method for the inverse synthetic aperture radar, which is based on the uniform rotation of a target in a fixed plane. To a maneuvering aircraft, the rotational velocity and the axis of a maneuvering target often vary with time. The traditional imaging method will blur the image of the target, and the target cannot be recognized. The available imaging algorithm for maneuvering targets can deal with the targets only with loworder Doppler variations of scatterers. The raw radar data show that the Doppler variations of scatterers are sometimes complicated for maneuvering targets. In this paper, an imaging algorithm for maneuvering targets based on adaptive chirplet decomposition is proposed. Without much more calculation, the novel algorithm can be used to deal with maneuvering targets with complex Doppler variations, as is verified with raw radar data.