This paper deals with the finite-time consensus control problem for a class of nonlinear strict-feedback multi-agent systems with heterogeneous dynamics. Due to the existence of unknown nonlinear dynamics of the system, this paper adopts the excellent approximation ability of fuzzy logic systems to design the consensus control protocol. Moreover, fuzzy logic systems and adaptive method are used to approximate the unknown system dynamics in the design of virtual controllers for different orders of the subsystems. The proposed direct adaptive fuzzy tracking controller for every agent is constructed via a backstepping design process. The proposed adaptive fuzzy controller can ensure that the outputs of all agents can track a common desired trajectory in finite time. Finally, the performance of the proposed fuzzy adaptive consensus control algorithm is demonstrated by two numerical studies of MASs including both homogeneous and heterogeneous dynamics of the interacting agents.