To solve the problem of poor tracking accuracy and slow response of intelligent car steering systems, a sliding mode control method of intelligent car steering system based on machine vision was proposed. Firstly, the intelligent car, which senses road information, was the research object. The principal analysis and mathematical modeling of the steering system were carried out. By analyzing the working process of the permanent-magnet direct current motor, the armature voltage balance equation and torque balance equation were listed, the transfer functions were obtained, and the transmission parameters were identified by experimental methods. Secondly, the sliding mode controller was designed for the above-mentioned controlled object, and its stability was proven. Matlab/Simulink simulation results verify the effectiveness of the sliding mode control algorithm. Finally, the combination of theoretical results and experimental verification shows that compared with PID（proportional integral differential）control, sliding mode control has better tracking accuracy and response speed at 20 Hz, the tracking accuracy is increased by 89.3%, and the system uncertainty can be overcome, especially for nonlinear systems.