To investigate the degradation mechanism of reinforced gravel subgrade under coupled traffic cyclic loading and fine particle contamination, and to determine the optimal placement depth of geogrids and critical threshold for fine particle contamination, a multi-stage dynamic loading model test was conducted using the DJM-500 electro-hydraulic servo loading system. Testing methods included micro-electro-mechanical systems (MEMS) wireless attitude sensors and dynamic soil pressure boxes. Surface settlement, soil pressure distribution, and coarse particle deflection characteristics of the gravel subgrade were systematically analyzed under different fine particle inclusion rates (0,10%,20%,30%). Results show that when fine particle contamination exceeds 10%, the subgrade enters an accelerated degradation phase. Under 30% contamination conditions, the final settlement of unreinforced and reinforced subgrades increase by 19.98% and 18.09% respectively compared to uncontaminated conditions. Geogrids significantly improve subgrade stability through stress diffusion and particle interlocking effects, reducing settlement by up to 33.35% and decreasing cumulative coarse particle deflection angle by 68.71%. Under uncontaminated conditions, soil pressure at the soil-rock interface decreases by 23.6%. However, fine particle contamination weakens the reinforcement effect of geogrids, with the difference in soil pressure between reinforced and unreinforced conditions decreasing by 75.66% at 30% contamination. Cumulative settlement of the gravel subgrade shows quadratic relationship with load cycles. The experiment determines the optimal placement depth of geogrids in 200 mm-thick gravel subgrades as 0.5B (B is the loading plate width), providing quantitative support for long-term deformation prediction of subgrades.