The aim of this article is to study a fuzzy-based decentralized adaptive control strategy for the non-strict-feedback fractional-order nonlinear large-scale systems with unknown control directions. In each step of the recursive processes, the fuzzy logic systems (FLSs) are employed to identify unknown nonlinear functions. To handle the difficulties caused by unknown control directions, a Nussbaum function technique is adopted. Furthermore, by introducing the dynamic surface control (DSC) technique into the adaptive backstepping recursive design algorithm, a fuzzy-based decentralized adaptive control strategy is formulated. Both the stability of the controlled system and the convergence of the tracking errors are proved by constructing the fractional-order Lyapunov functions. Finally, the validity and effectiveness of the designed decentralized control scheme are confirmed via two simulation examples.