Abstract: A sliding mode control method for rice transplanter path tracking based on interference observer is proposed to address the problems that the vehicle is prone to slip under unknown external disturbance during transplanter operation due to complex farmland environment and spatial variability, thus affecting the path tracking effect and long path tracking convergence time. Firstly, an extended kinematic model including slip is established, based on which the composite disturbances in lateral and heading directions are introduced, and a nonlinear disturbance observer is designed to accurately observe the disturbances and reduce the jitter of the sliding mode control. In order to avoid the singularity phenomenon and improve the convergence speed, a controller designed with fast non-singular terminal sliding mode surface and variable exponential power convergence law is proposed to weaken the disturbance. A comparative simulation of the interference-free model is established on the MATLAB software platform. Compared with the fast power convergence law, the variable exponential power convergence law selected in this paper is faster, and compared with the fast terminal sliding mode, the fast non-singular terminal sliding mode has better dynamic performance; the simulation results show that the nonlinear interference observer can accurately observe the composite interference; compared with the sliding mode controller without interference observer, this paper Compared with the sliding mode controller without interference observer, this paper can converge quickly and effectively, weaken the interference and reduce the jitter. In the actual paddy field operation test with the speed of 0.8 m/s, the average absolute deviation of the overall linear path tracking is 2.50 cm and the root mean square is 2.90 cm, and the overall deviation value is small, which can effectively improve the stability and speed of the control system and meet the path tracking accuracy.