The decision tree is an illustrious classification technique used for the pre-diction of the future data based on past experience. The decision tree is constructed using three major steps, first – constructing the decision tree to classify the data, second – pruning the decision tree to improve statistic certainty, third process the pruned decision tree to improve intelligibility. In this paper, we focus on second steps – pruning. Two famous algorithms are used for creating a decision tree, i.e. C4.5 and fuzzy C4.5. It presents the comparison of five well-known pruning methods. The performance of different pruning algorithm is evaluated using two main criteria size and classification errors after pruning decision trees. Different pruning methods have different impacts on decision tree.