Speech processing considered one of the most crucial areas of signal processing. Human speech is a natural form of communication, speech contains many information that can be used for determine the gender and identify of the person, thus we refer to biometrics technology, where speaker voice is considered one of the biometrics can be used for distinguished purpose. This paper presented simple speaker identification approach based on curvelet transform as a feature extraction technique and Backpropagation neural network BP NN as a classifier.