An image-restoration method which applies the iterative method to solve simultaneous linear equations is described. The advantages of this method are that the memory capacity to be used is minimal, the computation time is very short, and the man-machine interaction in the course of processing is easily effected. Owing to these advantages, this method seems to be superior form a practical viewpoint to other recently proposed linear-algebraic approaches for image restoration. The mathematical basis of this iterative image-restoration method is described and the suitability of this method is presented. The characteristics of this method are clarified.