Abstract: In order to compensate for the limitation of minimum variance (MV) beamforming in the aspects of robustness and contrast, an ultrasound imaging method based on eigenspace-based forward-backward minimum variance beamforming is presented in this paper. First, forward-backward (FB) spatial averaging, instead of the conventional forward-only spatial averaging is used to obtain a more accurate covarianee matrix; and then the calculated optimum weight vector is projected onto a signal subspace constructed from the eigenstructure of the covariance matrix; in the end, the obtained vector and the aperture data are calculated to obtain echo data. The experiments based on point pattern and speckle pattern are used to verify the proposed method. The experimental results show that the proposed method is less dependent on the choice of the amount of diagonal loading and enhances the contrast and robustness while the high resolution of the MV beamforming is retained.