Biometrics based person authentication methods are getting wide acceptance and they are replacing passwords because they are unique and secure. Palm print based person identification method is a technique which is more secure compared to other biometrics technique like finger print recognition, iris recognition and face recognition. Palm print recognition provides a unique identification of the people. Palm print identification method compares the input palm images with the images in the database. If the features of the sample palm images are matched with the features of the image of the palm in the database then it is considered as positive. In this proposed method, left and right palm print images of an individual is taken in six different orientations are taken. The orientation features are extracted using Gabor filter because it has multi resolution and multi orientation features. The orientation of the filter which gives the maximum filter response with the palm print is taken as the most dominant orientation of the palm print. Artificial neural network is used for classification because neural networks have the ability to learn complex input-output relationships. The learning process involve updating network architecture and connection weights so that the network can perform a specific classification efficiently. The back propagation neural network is used here. The back propagation is a multi-layer feed forward, learning network based on gradient descent learning rule. The back propagation neural network gives various performance analysis, error histogram, training state, confusion matrix and receiver operating characteristics graphs. This method provides better accuracy.