Abstract  Fuzzy neural controllers have the advantages of ease for knowledge expression and the ability of self-learning, and are able to control adaptively by updating the fuzzy rules and the membership functions. Nevertheless, the long training time usually discourages their practical applications in industry and the parameters over-updating may make system oscillate extensively. In this paper, a new strategy for optimizing the parameters updating algorithm of fuzzy neural controller is proposed. The only effect of parameters which affects the control performance significantly are updated. Also, based on fuzzy inference, the updating step is adjusted adaptively in accordance with the error and the change of error of the system. Two examples are simulated in order to conform the effectiveness and applicability of the strategy proposed in this paper.