Objective To establish a stable mouse model of mild-to-moderate closed head injury and evaluate its effectiveness. Methods A total of 170 SPF grade ICR mice were divided into a sham group (n= 34) and TBI group (n= 136). The TBI group was further divided into four subgroups, TBI-A, TBI-B, TBI-C, and TBI-D, with a drop height of 20 cm and drop weights of 60 g, 80 g, 100 g, and 120 g, respectively. According to the above conditions, the brain injury percussion was performed on the mice, 34 mice in each group. After modeling, the mNSS, fatigue rotating test, and Morris water maze test were used to assess balance and learning abilities of the mice. Brain water content analysis, evens blue staining, HE staining, GFAP immunohistochemistry, and TUNEL immunofluorescence were used to analyze pathological changes in brain tissue. Results Compared with the sham group, the mNSS and water maze escape latency in the TBI group increased to various degrees, whereas continuous walking time on the ratating rod decreased significantly, indicating severe neurological dysfunction in the TBI group. Brain water content and evens blue staining in brain tissue of TBI group mice were significantly higher than those in the sham group, suggesting blood-brain barrier damage and brain edema. Histopathological examination showed that neurons in brain tissue of the TBI group showed obvious degeneration and contraction, astrocyte proliferation, and the proportion of apoptotic cells was increased significantly. The damage degree increased with the increase in the weight of the impact. Comprehensive analysis revealed that weights of 60 g or 80 g with a height of injury at 20 cm simulated mild closed head injury in mice, whereas weights of 100 g or 120 g simulated moderate closed head injury in mice. Conclusions A stable mouse model of mild-to-moderate closed head injury was established using the principle of free fall, which lays the foundation to study the mechanism and treatment strategies for mild-tomoderate traumatic brain injury.