This paper presents an adaptive fuzzy control scheme for a class of nonstrict-feedback nonlinear systems with dead zone outputs and prescribed performance. By utilizing the monotonically increasing property of system bounding functions and the Nussbaum function, the design difficulties caused by the nonstrict-feedback structure and dead zone output are overcome. Combining backstepping technique with prescribed performance algorithm, a feasible adaptive fuzzy controller is designed to guarantee the boundedness of all signals of the closed-loop system and the prescribed tracking performance of the system. Finally, simulation results are depicted to illustrate the effectiveness of the proposed control approach. K E Y W O R D Sadaptive fuzzy control, backstepping technique, dead zone outputs, prescribed performance INTRODUCTIONBy utilizing fuzzy logic systems 1-4 or neural networks 5 to approximate the unknown nonlinear system functions, approximation-based adaptive fuzzy or neural control has received considerable attention in the past decades. In References 6-12, combining with backstepping technique, adaptive fuzzy or neural control was considered for strict-feedback nonlinear systems. In References 13-17, the authors extended the controlled nonlinear systems from deterministic systems to the uncertain stochastic nonlinear systems. However, the results obtained in References 6-17 cannot be directly applied to nonstrict-feedback nonlinear systems. For the strict-feedback systems, the nonlinear functions of the jth subsystem are independent of the latter state variables x i (j + 1 ≤ i ≤ n), and in the nonstrict-feedback systems, each subsystem function contains the whole state variables, which is the major difference between a strict-feedback nonlinear system and a nonstrict-feedback one. If the traditional backstepping control scheme is adopted in a nonstrict-feedback system, virtual control signal designed in the approach is the function of whole state vector, which makes the controller design difficult and leads to algebraic loop problem.Recently, many researchers have paid attention to adaptive fuzzy or neural control of nonstrict-feedback nonlinear systems. In Reference 18, the problem of observer-based adaptive neural output-feedback control was considered for stochastic nonstrict-feedback nonlinear system. In References 19-24, adaptive fuzzy or neural control was extended to nonstrict-feedback nonlinear systems with input nonlinearities such as time delays, 19 dead zones, 20,21 input saturation 22,23