This paper presents a robust H_∞ output-feedback control strategy for the path following of autonomous ground vehicles (AGVs). Considering the vehicle lateral velocity is usually hard to measure with low cost sensors, a robust H_∞ static output-feedback controller based on a mixed genetic algorithms (GA)/linear matrix inequality (LMI) approach is proposed to realize the path following without the information of the lateral velocity. The proposed controller is robust to the external disturbances and the uncertainties of the vehicle/environment parameters/states, including the tire cornering stiffness, vehicle longitudinal velocity, yaw rate and road curvature. Simulation results based on the CarSim-Simulink joint platform using a high-fidelity and full-car model have verified the effectiveness of the proposed control approach.