Objective To observe the clinical features and visual prognostic factors of ethambutol-induced optic neuropathy (EON). Methods A cohort study. Twenty-four inpatients (46 eyes) identified as EON in Neuro-Ophthalmology Department of Chinese PLA General Hospital from January 2014 to December 2017 were enrolled, including 14 males (26 eyes) and 10 females (20 eyes) with a ratio of 1.4/1 male/female. The average age was 42.79±15.12 years and the average weight was 62.46±12.31 kg. The average time duration between oral administration of ethambutol and occurrence of EON was 9.94±16.49 months. The average time of ethambutol duration was 7.06±11.68 months, with an average accumulative dose of 156.7±1 779.0 g and the average daily dose of 15.07±8.95 mg/(kg·d). All patients were tested with visual acuity, fundus photos, colour vision, OCT, visual field, VEP, orbital MRI and the gene of OPA1 and mitochondrial deoxyribonucleic acid(mtDNA). All the patients accepted drug withdrawal immediately after diagnosis, and were given the treatment of systemic nerve nutrition and improvement of microcirculation for 2 weeks. The time of follow-up was more than 12 months. According to whether the visual acuity (VA) in any of eyes was over than 0.1 at the last follow- up, all the patients were divided into two groups: the bad VA group (VA less than or equal to 0.1) and the better VA group (VA over than 0.1) group. The χ2 test and Fisher's exact probabilistic method test were used to compare the counting data between groups, and the Wlincox rank sum test was used to compare the measurement data. Multiple factors of VA outcome between the patients with bad or better va were analyzed by logistic regression. Results Thirty eyes (65.2%) had VA less than or equal to 0.1 and 5 eyes (10.9%) had VA over than 0.5 at EON onset. The VA of the rest 11 eyes (23.9%) was higher than 0.1 and lower than 0.5. At the last follow- up, 20 eyes (43.5%) had VA less than or equal to 0.1 and 9 eyes (19.6%) had VA over than 0.5, the VA of the rest 17 eyes (36.9%) was higher than 0.1 and lower than 0.5. Fundus examination revealed 7 eyes (15.3%) with optic disc edema. OCT revealed significant loss of the retinal nerve fiber layer (RNFL) in the affected eyes, mainly in the temporal RNFL of the optic disc. All patients had dyschromasia, mainly in distinguishing the color of red and green. The types of visual field defect was as following: central dark spot (52.2%), diffuse visual acuity decreased (30.4%), temporal hemianopsia (17.4%). Orbital MRI revealed that 12/24 (50.0%) patients had T2 lesions with T1 enhancement in 6/24 patients (25.0%). Genetic test showed that 4 patients (16.7%) had gene mutation. Among them, there were 2 patients with OPA1 mutation, 1 with mtDNA 14340 point mutation and 1 with the mtDNA 11778 point mutation. Thirteen patients showed better VA outcomes (over than 0.1) while 11 showed bad VA outcomes after discontinuation of ethambutol. Between the better VA group and the bad VA group, there were statistically significant differences in the daily dose of ethambutol and gene mutation(P=0.031, 0.023). The daily dose was related to visual prognosis of EON while only the daily dose of more than 18 mg/(kg·d) may lead to bad VA outcomes according to the logistic analysis (95% CI 0.007-0.736, OR=0.069, P=0.027). Conclusions EON may have OPA1 and mtDNA mutation with more bilateral eyes involved and less optic edema, which about 43.5% of the patients showed irreversible visual impact. The daily dose of ethambutol is related to the vision recovery. Key words: Ethambutol/adverse effects; Optic nerve diseases; Disease attributes; Prognosis; Root cause analysis