Abstract: Maximum a posteriori 3D restoration algorithm(3D MPMAP) with Markov constraint for three-dimensional optical-sectioning microscopy is proposed.According to the 3D feature of the microscopic image,3D point-spread-function is made,2D neighborhood is continued to 3D and the regularization parameter is simplified.As a result,the restoration of 3D microscopic image is achieved.Experimental results showed that the disturbance from out-of-focus signals is suppressed greatly in simulated microscopic images with different signal-to-noise ratio,and the spectra of the restored images are recovered greatly,the definition is improved obviously,and better effect in image restoration is reached for the microscopic images of actual biological specimens.