Abstract: Aiming at the engineering problem that fracture and instability of deep thick and hard roofs in China are prone to induce rock bursts, as well as the deficiencies in existing research including insufficient targeted design of hydraulic fracturing parameters, unclear dominant controlling factors of fracture propagation, and imperfect effect evaluation system, this study takes the 401102 working face of Mengcun Coal Mine as the engineering background to carry out research on the propagation law of hydraulic fractures in thick and hard roofs and the evaluation of rock burst prevention effect. Based on the key stratum theory, the medium-position thick and hard coarse-grained sandstone 57.98 m above the coal seam is identified as the main hazard-inducing stratum. The Petrel-Kinetix numerical simulation and L₁₆(4⁵) orthogonal test were adopted to analyze the influence law of six factors (including injection rate, fluid volume, and fracturing fluid type) on fracture propagation. Combined with the combined weighting-TOPSIS method, the optimal fracturing construction parameters for the target stratum were obtained through optimization. Meanwhile, the significance of the influence of each parameter on fracture propagation was quantified via multi-factor analysis of variance, and the fracturing fluid type was clarified as the dominant significant factor controlling fracture propagation. A comprehensive evaluation methodology for thick-hard roof hydraulic fracturing was established by integrating dynamic construction-condition monitoring, water-yield observation, surface and underground microseismic monitoring, and borehole televiewer verification. The optimized fracturing scheme achieves effective modification of the target horizon. Surface microseismic monitoring indicates that single-stage fractures exhibit an average horizontal length of 270 m, a bandwidth of 80 m, and a vertical height of 50 m. After fracturing, the rock burst risk of the working face is reduced, the daily released energy of microseismic events and the time proportion of high energy microseismic events are decreased significantly. The engineering input-output ratio reaches 1:2.16, and no secondary disaster caused by excessive fracturing occurs. This study can provide a theoretical basis and engineering reference for rock burst prevention and control in working faces with thick and hard roofs.