This study develops a fuzzy sliding mode controller (FSMC) based on a variable boundary layer. A fuzzy inference mechanism is used to on-line tune the thickness of the boundary layers of the controller. Minimum rule base has been used for the fuzzy inference system which results in low calculation effort. The aim of this paper is to design a controller which will eliminate the chattering of FSMC while maintains the robustness of the controller. To prove the effectiveness of this method, an experiment has been done using MATLAB/SIMULINK. In this experiment, the results of FSMC with auto-adjustable boundary layers are compared with the results of FSMC with fixed boundary layers and the results of FSMC with sign function. The results of the experiment confirm that the performance of fuzzy sliding mode controller based on auto-adjustable boundary layer method is superior to the fixed boundary layer method.