Abstract: This paper investigates the problem of motion parameter estimation in multichannel synthetic aperture radar and ground moving target indication (SAR-GMTI) system. A nonparametric, iterative adaptive approach based velocity estimation (IAA-VE) algorithm is presented. The core idea of this algorithm converts motion parameter estimation in the SAR system into solving an amplitude and phase estimation problem in array processing. Specifically, the array manifold of the radial velocity is constructed for the potential moving targets, and a weighted least square cost function is employed to estimate the complex amplitudes and velocities of moving targets with iteration metrics. Simulation results show that the proposed IAA-VE method can estimate the radial velocities of moving targets accurately and efficiently. Moreover, IAA-VE is shown to outperform a number of existing approaches.