This paper proposed a new classification method based on the discrete wavelet transform (DWT) combined with an automated classification mechanism based on adaptive network fuzzy inference system (ANFIS) for power transformer differential protection to discriminate between internal faults and no fault condition ( inrush condition) in three phase power transformers. For the evaluation of the developed algorithm, transformer modeling and simulation of fault and no fault condition are carried using Matlab/Simulink software Package. For each candidate internal fault or inrush current conditions current waveform suitable features are extracted by employing DWT. Then, a successfully trained adaptive network fuzzy inference system based classifier, developed utilizing inputs comprising the features extracted from a training set of waveforms is implemented for a testing set of sample waveforms. The simulation results obtained show that the method is faster, more reliable and accurate when compared with some of published research works in the area. Keywords: power transformer, inrush current, differential protection, discrete wavelet transform, adaptive network fuzzy inference system.