Abstract. Tracked vehicle platforms are widely employed in different areas where high levels of mobility are required such as agriculture and military due to their high traction efficiency in rough and moist environments. Automatic steering of a head feed combine harvester with crawler tracks that simultaneously modulates the speed and direction of the vehicle requires precise control of relative angular velocities of the left and right tracks and coupled operation of clutch and brake. This paper describes the development of a tracked vehicle kinematic model-based controller for an auto-guided head feed combine harvester that enhance its capability to follow along predefined a straight working line. A variable-speed hydrostatic transmission (HST) was employed to provide the functions of clutch and brake of the tested combine harvester. The path tracking and steering algorithms were implemented using a fuzzy logic method that considers the non-linear characteristics of the HST mechanism and various steering variables, such as angular velocity, lateral deviation and heading error. Prior to field testing, a feasibility of using the developed steering control algorithms was investigated in a simulation environment developed in our previous research. A rice harvesting test was conducted with an autonomous combine harvester equipped with the developed steering controller in a paddy field. The validity of the developed simulation was confirmed by comparing the tracking trajectory with field test in same condition. The autonomous combine harvester followed the predefined line path with acceptable tracking, showing lateral error and heading error within 15cm and 3deg respectively.