The intelligently tracked vehicle is fit for working on soft ground, and its auto drive function is of great significance. This paper introduces the path following system of an intelligently tracked vehicle using fuzzy neural networks. The system, which is an integral combination of the fuzzy system and artificial neural system, can not only deal with information expressed linguistically, but also possess the same function of supervised learning and knowledge acquisition as the artificial neural networks. Therefore, the system is more adaptable to its working environment. The test shows that the system follows the changing path at a rapid speed, and can efficiently satisfy the demands of real time path following.