This paper presents annew approach to intelligent gear fault diagnosis in the field of rotating machinery condition monitoring.The prediction process is focused on the vibration signals collected from the accelerometer sensor mounted on the test rig for critical components monitoring. The pre-processed signals were decomposed into several signals containing one approximation and some details using Wavelet Packet Decomposition and thestatistical features were extracted from 4th level WP decomposition. The J48 algorithm which uses gain ratio to determine the splits of decision tree and to select the most important features were used for selecting the predominant features and selected features were fed as input for training and testing of Multi class Least Squares Support Vector Machines (MCLS-SVM) together with different kernel functions such as radial basis function kernel, linear-kernel, Polynomial-kernel functions for classification and predict the fault condition of the components and machines.Analysis is done for various LS-SVM models such as One against One and One against All. The proposed approach is applied to fault diagnosis of gear, and testing resultsshows the proposed approach can reliably recognise different fault categories and severities. © Research India Publications.