Post-fracturing flowback significantly influences the evaluation of fracturing efficacy, with proppant backflow during this stage substantially affecting later production stages. At present, the control of proppant backflow relies largely on empirical methods used at production sites, indicating a lack of theoretical research. This study employs an experimental setup designed to simulate proppant backflow using artificially created fractures in rock slabs, exploring the influence of rock properties, closure pressures, and proppant characteristics. The analysis indicates that (1) under low closure pressures, the influence of fracture roughness on proppant backflow surpasses that of the closure pressures; (2) there is a positive correlation between fiber concentration, proppant size, and the critical flowback rate, while sand concentration and fluid viscosity correlate negatively; (3) increased fracture roughness accelerates proppant and fluid flow within the fractures, complicating flow trajectories and favoring the formation of highly permeable channels. Recommendations for field fracturing include using a tiered sanding approach with 30/50, 40/70, and 70/140 mesh sizes supplemented by fiber at the end, which effectively prevents proppant backflow. Lowering fracturing fluid viscosity through gel breaking after fracturing, during the gradual closure phase, can prevent proppant backflow; adjusting operational parameters in high-yield areas to form high-speed flow channels is also suggested.