Thanks to many advantages of the iris, such as stable, nonintrusive, and unchangeable, iris recognition has been a research hot-subject in the biometrics identification field. However, the abundant textures and complex structures of irises lead to a great number of troubles for feature extraction and encoding. In order to simplify the feature extraction and encoding method and improve the efficiency of iris recognition, an iris block-encoding method based on statistic of local information is proposed. Firstly, it achieves preprocessing of the original eye image, such as iris localization, and gets the normalized iris features image. Secondly, according to the comparative relationships between local information and global information, local information and local information, it accomplishes the iris block-encoding. Thirdly, it calculates the Hamming distance between different iris codes and obtains the recognition result according to Hamming distance. Experimental results demonstrated that the proposed method is effective and feasible. It also has high recognition accuracy and speed.