:To address the problem of reduced positioning accuracy in hilly terrain tractors when GPS/INS combined positioning systems experience unstable satellite signals, this study proposes a pose estimation method based on Random Sample Consensus (RANSAC) plane fitting using LiDAR point cloud data. The algorithm"s pose estimation principle is analyzed, and an error model for pose calculation is established. Compared to the ICP method, it effectively reduces the number of iterations. This innovative approach achieves efficient processing of large-scale point cloud data and precise determination of tractor pose in complex terrain. After collecting environmental point cloud data of tractors in various postures via LiDAR, the RANSAC algorithm is applied to remove outliers from the point cloud data. By fitting the ground plane equation to the point cloud data, the tractor posture in hilly terrain is solved. Experimental results demonstrate: The roll angle of -0.8° and pitch angle of -0.3° obtained from the horizontally placed LiDAR posture solution validate the feasibility of the RANSAC algorithm. In subsequent standard slope tests, the calculated roll angle of 11.167° for a tractor on a single-sided bridge in hilly terrain matched the actual value of 11.71°, achieving 95.36% accuracy. The pitch angle calculation on a 30% standard slope yielded 14.354°, achieving 86.23% accuracy. This method effectively addresses the challenge of attitude sensing for tractors in hilly terrain when satellite signals are unavailable.