Edge-preserving blind image restoration using a hierarchical Bayesian model is proposed in this paper. The restoration problem, when the point-spread function (PSF) of the degradation system is partially known, is examined. The simultaneously autoregressive image model and the fixed-f covariance model are used under a Bayesian framework. The evidence analysis approach is then used to simultaneously estimate the parameters and the image iteratively. An edge-preserving regularization operator is utilized to carry out regularization according to the local image characteristics. Numerical experiments and conclusions are given to show the effectiveness of our method.