In this study, it was aimed that making epilepsy diagnosis by automatically evaluation of EEG records. Diagnosis system consists two steps which are feature extraction/selection and classification. Discrete Wavelet Transform (DWT) and Artificial Neural Networks (ANN) were used to determine attribute vectors and classification, respectively. Classification accuracy was achieved as 99.62% by examining effects of varied wavelets on Multi Layer Perceptron (MLP) networks which have different architecture and were trained different learning algorithms.