BACKGROUND: Hemifacial spasm (HFS) is characterized by progressive, paroxysmal, and involuntary convulsions on one side of the face. We have conducted in-depth exploration on the puncture approach through the mandibular angle, which is an important supplement to the first 2 approaches (i.e., premastoid approach and the postmastoid approach), especially for patients who cannot find a suitable way before and after the mastoid process. OBJECTIVES: To investigate the effect of computed tomography (CT)-guided percutaneous mandibular angle radiofrequency thermocoagulation (RFT) of facial nerve through the stylomastoid foramen in treating HFS. STUDY DESIGN: A retrospective, observational study. SETTING: Pain Department, Jiaxing and Hangzhou, China. METHODS: A total of 89 patients with HFS who underwent CT-guided RFT in the Pain Department of Zhejiang Integrated Traditional Chinese and Western Medicine Hospital and the Pain Department of Jiaxing First Hospital, from June 2020 to June 2022, were retrospectively analyzed, including 29 men and 60 women, aged 34~88 (59.8 ± 11.1). They were divided into 3 groups: anterior mastoid approach group (Group A, n = 38), posterior mastoid approach group (Group P, n = 26), and mandibular angle approach group (Group M, n = 25), according to the different puncture approaches. Puncture time, minimum stimulating current triggering facial muscle twitches, temperature at the end of RFT and duration time of RFT at this temperature, and total treatment time, as well as degree of facial paralysis and complications one-day postoperation, were compared among the 3 groups. RESULTS: The puncture times (mean ± SD) of Group A, Group P, and Group M were (30.63 ± 4.88), (31.35 ± 5.89), and (35.08 ± 5.76), respectively, and the differences were statistically significant (P = 0.006). The puncture time of Group M was longer than that of Groups A and P (P < 0.05). The minimum stimulating current triggering facial muscle twitches in the 3 groups were (0.49 ± 0.16), (0.43 ± 0.14), and (0.28 ± 0.09), respectively, with a statistically significant difference (P = 0.000). The minimum stimulation current in Group M was less than that in Groups A and P (P < 0.05). The temperature at the end of RFT of the 3 groups was (78.29 ± 7.91), (76.54 ± 8.10), and (66.60 ± 8.00), respectively, and the differences were statistically significant (P < 0.001). The temperature of Group M was lower than Groups A and P (P < 0.05). There were no significant differences among the 3 groups in the total operation time or the degree of facial paralysis one-day postoperation (P > 0.05). No hematoma, infection, hearing impairment, or other complications were reported in any patients. LIMITATIONS: The nonrandomized nature, small sample size, and retrospective design are limitations of this study. CONCLUSIONS: CT-guided RFT through the stylomastoid foramen is an effective treatment of HFS. Compared to the poster and anterior mastoid approaches, the mandibular angle approach was found to be more effective in terms of reduced minimum stimulating current and reduced-end RFT temperature, which means fewer potential complications to the patient postsurgery. KEY WORDS: Hemifacial spasm, radiofrequency thermocoagulation, stylomastoid foramen, CT-guided