Circular ground-based synthetic aperture radar (GBSAR) is a new type of GBSAR with three-dimensional (3D) imaging capability. Its relatively low bandwidth-to-frequency ratio and donut shaped spectrum cause high sidelobes problem, which seriously affects the quality of SAR images. To solve this problem, a multi-phase-center sidelobe suppression method for circular GBSAR based on sparse spectrum is proposed in this paper. This method optimizes the sparse spectrum distribution of multi-phase-center circular GBSAR to suppress the originally high sidelobes. In this paper, the signal model of multi-phase-center circular GBSAR is established, the point spread function (PSF) is derived, and the relationship between the sparse spectrum distribution and the sidelobe level is also analyzed. Moreover, we provide the optimal distribution of the sparse spectrum, thus the optimal distribution of multiple phase centers can be calculated with given imaging geometry, providing the best sidelobe performance. The proposed method is verified via simulation experiments.