The problem of using neural networks for military vehicle classification on the basis of ground vibration is presented in this paper. One of the main elements of the system is unit called geophone. This unit allows to measure ground vibrations in each direction for certain period of time. The value of amplitude is used to fix LPC values of each vehicle. Because the multilayer perceptron is used, the learning set has to be prepared. Please find attached the results of using neural network such as: example of learning, validation and test sets, structure of the networks and learning algorithm, learning and testing results.