This paper presents a robust control methodology based on a disturbance observer and an optimal states feedback for Takagi–Sugeno fuzzy system. Firstly, the nonlinear systems were solved by applying a sector nonlinearity method to get the inner linear subsystems and outer fuzzy membership functions, which guaranteed the conversion without any loss generality characteristics of the system. Secondly, an exponentially convergent disturbance observer was constructed to the system with an assumption that the system states are temporarily bounded. Thirdly, a states observer was built by poles placement of linear quadratic regulation optimization, which was used to place the system states error poles located on the stable region. Finally, simulation examples were given to figure out that the proposed controller is effective to control the Takagi–Sugeno fuzzy system. The obtained results are disturbance mostly rejected, state estimation errors are quite small, and the output signal precisely tracked input signal.