Experimental results on an automatic lane keeping control study are presented in this paper. A magnetic reference system was used to provide the vehicle lateral tracking error as well as future road curvature information and vehicle speed measurement to the vehicle steering controller. The front wheels of the vehicle are steered according to vehicle lateral displacement, future road curvature, vehicle yaw rate and lateral acceleration, and vehicle speed. The control algorithm used to design the feedback and feedforward controllers is known as the Frequency-Shaped-Linear-Quadratic Preview (FSLQ-preview) optimal control design method. The closed loop response of the vehicle was examined under a wide variety of test conditions, including low tire pressure, measurement noise, perturbed reference system, hard braking, and snowy road.