Based on the concepts of feedback linearization and backstepping approaches, this paper proposes a decoupled adaptive type-2 fuzzy controller (DAT2FC) scheme to treat the stabilization of nonlinear benchmark problem- TORA system. Firstly, partial feedback linearization was employed to simplify the nonlinear system to a simpler form. Therefore, the transformed system can be decoupled into two subsystems by the concept of backstepping. Two type-2 fuzzy controllers corresponding to the sub-systems are designed to follow the desired output. The corresponding adaptive laws are derived by the backpropagation learning algorithm. Therefore, the stabilizing controller can be obtained for the whole system. Finally, simulation and comparison results are shown to demonstrate the effectiveness of DAT2FC scheme.