Abstract: An adaptive fuzzy control approach is proposed for a class of multivariable nonlinear systems with disturbances and uncertainties. The multivariable systems are composed of m interconnected subsystems. The unknown functions in each interconnected subsystem are non-affine. It is difficult to control this class of systems due to the existence of the uncertain non-affine functions and the couplings among interconnected subsystems. This difficulty is overcome by using the mean value theorem, the fuzzy system, the backstepping design method and introducing Nussbaum-type functions. In addition, compared with most of the research results, the proposed approach reduces the quantity of the online adjusted parameters. The approach can achieve that all the signals of the closed-loop system are bounded. Simulation experiment demonstrates the effectiveness of the control approach.