Precise path tracking control of autonomous vehicles is an area which is beginning to benefit from an increased awareness of its importance and its challenge. One of the challenges comes about due to the need for increased accuracy and robustness. Kinematic models are primarily used for path tracking control due to their generic and more simplistic nature when compared to dynamic models. However, control based solely on the kinematic model rarely takes into consideration any of the dynamics that are present, which can degrade the performance. This paper looks at the path tracking control of vehicles, specifically agricultural tractors, incorporating compensation for inevitable steering dynamics. The path tracking control is achieved via a state transformation and full state feedback, equivalent to the backstepping algorithm. The backstepping approach with steering dynamics compensation is compared to more traditional kinematic model only approaches which assume such dynamics are insignificant when compared to the bandwidth of the path tracking controller itself. For comparison, such traditional approaches are implemented with both PD and 2-sliding mode controllers employed for low level steering control. Simulation results are carried out and show that incorporating the steering dynamics into the design can yield improvements in performance.