Intermittent monitoring of abnormal muscle response (iAMR) has been reported to be useful for improving the surgical outcome of microvascular decompression (MVD) for hemifacial spasm (HFS). However, iAMR has not elucidated the relationship between AMR change and the corresponding surgical procedure, or the pathogenesis of AMR and HFS. The purpose of this study is to clarify the usefulness of continuous AMR monitoring (cAMR) for improving the surgical results of MVD and for understanding the relationship between AMR change and corresponding surgical procedure, and the pathogenesis of AMR and HFS. Fifty consecutive patients with HFS treated by MVD under cAMR monitoring, which continuously records AMR every minute throughout the surgical period, were retrospectively analyzed. The patients were assessed for the presence of HFS 1 week after the surgery and at final follow-up. Forty-six patients showed the complete disappearance of HFS. In 32, AMR disappeared abruptly and simultaneously with decompression of an offending vessel. AMR showed dynamic and various changes including temporary disappearance, or sudden, gradual, or componential disappearance before and during the decompression procedure, and even during the dural and skin closure after the initial decompression procedure. Facial spasm remained in four patients despite permanent AMR disappearance. cAMR monitoring improves the outcome of MVD. Although the main cause of HFS and AMR is vascular compression at the facial nerve, hyperexcitability of the facial nucleus is also involved in the pathogenesis of HFS and AMR. The proportional involvement of these causes differs between patients.