Abstract: A type of direct adaptive control method based on generalized dynamic fuzzy neural networks for underwater vehicles was proposed in this paper. The proposed control method, which needs neither prior fuzzy neural networks structure knowledge nor prior training phase, could be used to build the underwater vehicles inverse-dynamic model through online adaptive learning algorithm. The underwater vehicles direct adaptive controller based on this kind fuzzy neural networks is proposed in this paper, and then the stability of the resulting underwater vehicles closed-loop control system is proved using Lyaponov stability theory. The validity of the proposed control method has been verified through computer simulation experiments using an underwater vehicle model.