High irregularity of Iris tissue is the main reason of its usage in identification systems. In some literatures, this irregularity has been known as a result of a stochastic genetic process. In this paper an efficient method based on the Poincaré section and modified version of fractal dimension is used. In fact, this method is introduced a pattern for decision making about series specification. Therefore, this criterion is utilized to show the stochastic behavior of Iris image. In this way, the assumption of existence of regularity in Iris pattern can be rejected. This method is applied to three areas in this paper using" CASIAI" and" CASIAIII (Interval)“databases. Obtained results demonstrate efficiency of this method.