Classification is the one of the well-known techniques in data mining. Based on the attributes of the object, classification assigns an object to one of numerous pre-defined categories. If information gain is not good then split attributes values into groups until we get better classification ratio. J48 is the one of the most frequently used classification techniques. In this paper, J48 is employed to effective prediction analysis of Iris data set. Three types of Iris flower with 250 instances and five attributes is used as test and training data. The results show that the accuracy of prediction is improved when compared with the existing ID3 method. © 2015 Elsevier B.V., All rights reserved.