Clinically, a consensus of the treatment strategies of ACL grade 2 injury is not reached. Therefore, the present study established a proprioceptive deficit model of the ACL via an arthroscope to aid with further studying ACL grade 2 injury and treatment strategies. 12 cynomolgus monkeys were randomly divided into the model group and the blank control group. In the model group, 1/4 of the ACL was cut under an arthroscope, whereas no intervention was performed in the blank control group. Physiological data including the maximum degree of knee flexion, the thigh circumstance and the calf circumference were measured, and the Pivot-shift, Anterior and Lachman tests were also performed. Moreover, electrophysiological data including somatosensory evoked potentials (SEPs) and motor nerve conduction velocity (MNCV) were measured. SEPs and MNCV were assessed for the latent period and amplitude. Comparing the data before and after the surgery, in the blank control group, no significant difference was observed. In the model group, significant difference was observed in the Pivot-shift, Anterior drawer and Lachman test (p < 0.05), indicating the instability of the ACL. Moreover, the latent period of SEPs and MNCV were significantly increased (p < 0.009), whereas the amplitude of SEPs and MNCV was significantly decreased (p < 0.009), indicating a degenerated proprioception of the knee joint. Following cutting 1/4 of the ACL, the knee joint became unstable and proprioception was declined. The results indicated that the proprioceptive deficit model of the ACL was successfully established, which could promote further studying ACL injury.