In this paper, an algorithm for enabling fast near-field multiple-input multiple-output synthetic aperture radar (MIMO-SAR) imaging is proposed and analyzed. The drastic reduction in the near-field MIMO-SAR scanning time is made possible by introducing high sparsity in the synthetic aperture. In addition, the proposed algorithm addresses the problem of high sidelobe levels, which results from random under-sampling. It is shown that the proposed approach significantly reduces the scanning time without sacrificing resolution. Moreover, the proposed approach manages to attenuate the sidelobe level significantly, which improves the dynamic range of the MIMO-SAR images.