We propose an iris localization algorithm based on gray distribution features of eye images.In this algorithm,we don't need to find all of the iris boundary points,and we should detect only three points on the iris inner and outer boundaries respectively.Then utilizing three points on the same boundary,we can calculate the circle parameters.We have done a large number of experiments on the CASIA iris image database,and the experimental results demonstrate that the localization results of proposed algorithm are more accurate and rapid than any other classical algorithms,such as Daugman's algorithm and the algorithm of edge-detection operator combining with Hough transform.