Abstract This study aimed to investigate telemedicine reliability and usability for facial dystonia. Eighty-two telemedicine recordings from 43 adults with blepharospasm (12,27.9%) and hemifacial spasm (31,72.1%) were obtained (mean age 64.5+9.3 years, 32 female (64.4%)). Two recorded in-hospital telemedicine visits were arranged on the same day as in-person visits at baseline and 4-6 weeks. The third non-recorded home-based telemedicine visit was held 2 days prior to the third in-person visit at 12-14 weeks. After 8 weeks, the neuro-ophthalmologists who performed the in-person visits also evaluated the telemedicine videos records. Spasm gradings by Jankovic Rating Scale (low, grades 0-2; high, grades 3-4), signs and symptoms indicating botulinum toxin complications were collected. Intra-rater agreements in assessing spasm gradings were moderate (severity: kappa=0.42, 95%CI 0.21 to 0.62, frequency: kappa=0.41, 95%CI 0.21 to 0.61) with substantial agreement in detecting lagophthalmos (kappa=0.61, 95%CI 0.36 to 0.86). Adding symptoms to signs increased sensitivity and negative predictive value in detecting lagophthalmos (66.7% to 100% and 94.3% to 100%) and drooping lips (37.5% to 75% and 93.6% to 96.4%), respectively. High mean usability score of 6.5(SD0.8) out of 7 was determined by “Thai version Telehealth Usability Questionnaire.” Therefore, telemedicine could be an alternative platform to evaluate facial dystonia.