<p>Rockfalls, a common type of sudden geological hazard, pose significant threats to human society due to their destructive power, wide-ranging impacts, and complex formation mechanisms. Existing numerical simulations of rockfalls often rely on idealized assumptions that fail to account for real-world phenomena, such as collision-induced fragmentation and air resistance, during the rockfall process, thereby compromising simulation reliability and prediction accuracy. To address these limitations, this study developed a high-precision rockfall dynamics simulation method that integrates physical modeling with multi-source data, utilizing Unreal Engine 5. The technique considers key controlling factors, including rock type, mechanical properties, underlying surface characteristics, and air resistance, to establish dynamic parameters for unstable rock masses. By employing a position-based dynamics algorithm and a graded fragmentation mechanism, this approach effectively simulates the complete dynamic behavior of unstable rock masses—ranging from instability initiation, collapse, and collision fragmentation to secondary ejection—while outputting critical parameters, including three-dimensional coordinates, velocity, kinetic energy, and movement trajectories. The methodology was applied to simulate the rockfall dynamics at Bijiashan, Chongqing. The results demonstrate that the distribution of collapsed rocks closely aligns with the historical deposition pattern at the slope base. The proposed high-precision rockfall dynamics simulation method significantly enhances numerical modeling capabilities for collapse disasters in complex environments, offering crucial theoretical foundations and practical value for disaster prevention and control.</p>