Abstract: Synthetic Aperture Radar(SAR) tomography is an important technique for target elevation information inversion and reconstruct the 3-D structure of the target via multi-pass observations. At present, SAR tomography is usually implemented based on the far-field assumption where the range between the radar and the target is far larger than the target size. Therefore, the look angle and the target scattering characteristics have good consistency. However, in the case of near-field case where both the look angle and the target scattering characteristics change seriously, the existing SAR tomography technique will fail. In order to solve the problem, this paper proposes a new method to reconstruct the 3-D structure of near-field target via compressive sensing MIMO SAR. The main contribution of this method involves: (1) Using the multi-strong point to compensate the amplitude and phase error between multiple channels to improve images quality; (2) Using scattering structure to extract scattering singular values and improve registration accuracy; (3) Using deramp method and 3-D geometric relationships to get the elevation information. Besides, we have successfully implemented near-field ground-based MIMO SAR tomography via compressive sensing. The presented approach has been evaluated via real Ku-band MIMO SAR data.