Abstract - Iris recognition system is one of the most reliable identification systems. However these systems are still uncommon because of many drawbacks for instance unreal time processing and inefficient identification methods. One of the critical tasks in iris recognition is detection of pupil. Pupil boundary detection method depended on the circular or elliptic shape of the pupil which could lead to inaccurate results of boundary detection. Thus, this can affect on the next step of iris recognition systems. Therefore, in this paper, we proposed a new method for pupil boundary detection. This method is based on thresholding the eye image by using the negative function. Next, the Four Neighbors method is used to detect the pupil boundary. Our method was tested on all CASIA V1.0 eye images and we attained 100% pupil boundary detection accuracy with acceptable computational time that can be implemented for real time processing.