Abstract: A systematic review was presented for the current research progress on tire modeling and experimental techniques in the off-road vehicle tire-soil interaction. Two specific tests were involved, including the soil mechanical properties and the indoor single-wheel soil tank test. The soil mechanical properties test was divided into the soil pressure, shear strength, and soil cone index measurement. The bevameter was generally used to determine the compressive and shear strength of soils in the tire-soil interaction, which were greatly varied in the fields. Three parts were mainly divided into the bevameter, including the loading, testing, and acquisition device. The less stable and less accurate operation was found for the original bevameter that was driven by a simple hydraulic system. The servo technology and automatic control system were utilized for the higher level of intelligence so far. The soil cone index was measured by the cone index instruments. Since the measurement accuracy of the cone index meter was directly related to the smoothness of the cone penetration speed, the improved cone index meter was changed from the initial human-driven to being driven by a constant-speed hydraulic cylinder or a servo motor, particularly for the simultaneous measurement of multiple physical parameters, such as soil moisture content, in order to calibrate each other and improve the measurement accuracy. Three approaches to tire modeling were summarized in the tire-soil interaction: empirical, numerical, and semi-empirical modeling. Among them, the empirical tire-soil model was confined to a single application scenario suitable for the rapid evaluation of tire performance. The reason was that the empirical equations in the model were excessively dependent on the parameters of tires and soils, indicating the lower computational demand, but the lower model accuracy in the empirical modeling. The numerical modeling was used to accurately simulate the force characteristics of tires in different soils, but the complex structure made it difficult in the real-time simulation. Some current simplified numerical models run in real time with the ever-increasing computing power of the computer. The semi-empirical modeling integrated the tire-soil interaction and the empirical fitting, indicating an excellent balance of complexity and accuracy. The embedded platform was easily applied for the real-time model in the off-road vehicle control strategy. At the same time, several application examples of each modeling were given for comparison. In conclusion, the future directions of tire modeling and test techniques in the tire-soil interaction were proposed to combine the latest progress in advanced fields, such as intelligent sensing, intelligent measurement, machine learning, and whole-vehicle collaboration. The finding can provide a strong theoretical basis and technical reference for the off-road vehicle tire-soil interaction.