The voice of each individual is unique because the fact that each person have a different size of vocal cord and the shape of the body. The length of the vocal cord decides the pitch and range of frequency of the voice of an individual. In this paper, we have made an attempt to use this unique feature of the individual for the purpose of identification and recognition. The proposed technique uses the concept of Fourier Transformation (FT) to extract the pitch of the voice and it also uses the method of noise removal and extraction of the noise free voice. The pitch for each individual is extracted and saved in the database. To recognize the voice, the proposed model makes use of neural network to train and classify the new voice.