Inverse synthetic aperture radar images maneuvering targets, and during the coherent processing interval time, imaging projection plane and the scale of cross-range change with time, so many parameters are difficult to accu- rately extract and more prior knowledge can't be acquired. In general, the robust rangcdoppler (RD) imaging algorithm is used. But the conventional RI)imaging algorithm is based on the hypothesis of the target rotating uniformly and sam- pling uniformity in azimuth. For maneuvering target imaging, RD algorithm will make the image fuzzy. Especially for ra- dar gapped data, the performance of imaging descends greatly and even cant be identified. This paper introduced a range instantaneous Doppler imaging algorithm based on sparse representation and time-frectuency transform which can effec- tively image the maneuvering target. hhe experimental results validate the effectiveness and feasibility of this approach.