Since hydrodynamic coefficients of autonomous underwater vehicle are varied with change of operating conditions which make it difficult to model it accurately, thus an adaptive fuzzy sliding mode control scheme is proposed for position tracking problem. Moreover, an innovative neural network compensator is designed to completely counteract effects of uncertainties and external disturbances. To diminish tracking error and enhance reaching time, three fuzzy logic systems are exploited; the first and second fuzzy logic systems are used to update slope of sliding surface and hitting control gain of sliding mode method, respectively. In addition, the third adaptive one is employed to estimate online unknown nonlinear functions of system dynamic as well as external disturbances. Finally, simulation results illustrated the effectiveness of the proposed approach.