Abstract: To investigate the influence of fine aggregates on the deformation characteristics of graded crushed stone, a modified prediction model was established based on revealing the effects of fines content （F_c）, fines-to-sand ratio （Dr）, and plasticity index （PI） on the resilient modulus. Taking limestone and basalt chips as research objects, repeated load triaxial tests were carried out to analyze the influence laws of F_c, Dr and PI on the resilient modulus. Five classical prediction models were compared, and the optimal one was selected via statistical indicators. Then F_c, Dr and PI were introduced as new variables to modify the optimal model, and the modified model was verified with measured data. Results show that when F_c is in the range of 1.5%~3%, the mixture exhibits optimal deformation resistance. The influence of Dr is strongly dependent on F_c: a lower Dr under low F_c or a higher Dr under high F_c both help increase the resilient modulus. The resilient modulus decreases linearly with increasing PI. The recommended thresholds for fine aggregate design parameters are F_c = 1.5%~3%, Dr = 0.18~0.28, and PI ≤ 8. The modified model achieves a prediction accuracy R² ＞ 0.8 under various fine aggregate conditions. This study demonstrates that the proposed model can accurately characterize the influence of fine aggregate parameters on resilient modulus, providing a theoretical basis and technical support for the design optimization of graded crushed stone materials.