Introduction/aimsTo facilitate further investigation into the mechanisms of facial nerve regeneration, a simple and reliable model of facial nerve crush injury is essential. Nevertheless, the establishment of such models lacks standardization and repeatability, while the healing capacity of the nerve is often overlooked, potentially affecting future studies.MethodsWe made facial nerve trunk crush injury models with different pressing times and detected the changes from the distal nerves to the motoneurons via behavior analysis, electrophysiological test, and histomorphometry analysis.ResultsIt revealed a particular capacity for self‐healing following facial nerve crush damage because there was almost no facial motoneuron apoptosis in the MC group during the observation period, and rats in MC group had total facial paralysis in behavioral tests following surgery and varying degrees of recovery 28 days postoperatively with no treatments. As the pressing time increased, the latency, wave amplitude, nerve fiber damage degree, nerve axon ratio, myelin thickness, electroneurograph (ENoG) value, ultrastructural damage, abnormal morphological changes, and the buccal muscle atrophy of each MC group gradually increased or got worse during the observation period. However, after 28 postoperative days, only the ENoG values of the M10min and M12min groups were beyond 90%, indicating no self‐healing.DiscussionIt suggests that a stable model of peripheral facial palsy may be created by applying a 12.5 cm mosquito clamped to the facial nerve trunk for at least 10 min, which laid the foundation for the subsequent research to objectively evaluate facial nerve regeneration.