Abstract: Effective rejection of the time-varying clutter originating from slowly moving vessels and surrounding tissues is very important for depicting hemodynamics in ultrasound color Doppler imaging. In this paper, a new adaptive clutter rejection method based on Hankel Singular Value Decomposition (Hankel-SVD) is presented for suppressing non-stationary clutter. In the proposed method, a Hankel data matrix is created for each slow-time ensemble. Then the orthogonal principle Hankel components can be obtained through the SVD of the Hankel data matrix. It achieves non-stationary clutter suppression by reconstructing the flow signal with only the high order principle Hankel components, which are estimated from the frequency content carried by the principle Hankel components. To assess its efficiency, the proposed Hankel-SVD based method is applied to synthetic slow-time data obtained from a Doppler flow model and carotid arterial complex baseband data acquired by a commercial ultrasound system (Sonix RP). The resulting flow and power images show that the proposed method outperforms the traditional IIR and polynomials regression filter in attenuation of high intense non-stationary clutter signal. It is also adaptive to highly spatially-varying tissue motion and can automatically select the order of the filter, which leads to improved distinguishing between blood and tissue regions compared to other eigen-based filters.