Abstract: Grouting reinforcement technology for fractured coal and rock masses is characterized by high concealment, complex procedures, and high construction costs. It is relatively difficult to design grouting schemes and study the grouting reinforcement mechanism through in-situ underground experiments. At present, there are few studies focusing on the design of grouting schemes, and the grouting reformation mechanism has not been fully revealed. To address these issues, a grouting reinforcement test system applicable in laboratory settings was developed for fractured coal and rock masses. Specimens of fractured gypsum-based rock masses with low dispersion characteristics were prepared, and the water–cement ratio of the cement slurry was used as the experimental variable. Grouting reinforcement tests under different water–cement ratios were carried out in the laboratory using the constructed system, and the grouting reformation mechanism was revealed. The results showed that: ① after grouting and reformation, the peak stress of the fractured gypsum-based specimens ranged from 3.64 to 5.21 MPa, indicating that their bearing capacity was effectively improved compared to the initial average residual stress of 3.11 MPa. ② With the increase in the water–cement ratio, the viscosity and flowability of the cement slurry continuously improved. Even fractures with a width of approximately 0.1 mm were effectively filled under high water–cement ratios. The structural framework effect of the reformed specimens continued to strengthen. When the water–cement ratio increased from 1.00 to 1.50, the average peak stress of the grouted and reformed specimens rose from 3.79 MPa to 5.09 MPa, representing an increase of 34.30%. ③ Within a certain range, slurry with a high water–cement ratio significantly enhanced the integrity and mechanical properties of the fractured rock mass. This clarified that the essence of grouting reinforcement lay in the filling of fractures by the slurry and the formation of the structural framework effect.