Purpose . To evaluate the efficiency and safety of three different techniques in the surgical treatment of epimacular fibrosis. Material and methods. A comparative evaluation of surgical treatment results was carried out in three groups of patients with epimacular fibrosis. In the first group – 20 patients (20 eyes), epiretinal membrane (ERM) was removed without vitrectomy. In the second group 30 patients (30 eyes), epiretinal membrane was removed after subtotal vitrectomy. In the third group – 30 patients (30 eyes), epiretinal membrane was removed after local vitrectomy. Local vitrectomy was performed by the following procedure: installing three 25G ports, induction of posterior hyaloids detachment in the macular area, local vitrectomy within the zone of vascular arcades (approximately 20% of the total vitreous volume), except for the area over the optic nerve head, layer-by-layer staining with MembraneBlueDual, removal of ERM, and then ILM. In all the groups a control of visual acuity was monitored, as well as intraocular pressure, the retinal thickness in the central zone, and the thickness of the retinal nerve fiber layer (RNFL) in different sectors. The OCT was used for a control the optic nerve head. The control of fusion of flickering frequency was made. The control of the optic nerve head was made using OCT and the control of fusion of flickering frequency (CFFF). Sensitivity of the macular zone of the retina was determined using the Maia microperimeter, and peripheral zones (Peripheral 60-4) were detected by means of the Humpfrey computer perimeter. The patients were examined before the surgery and 1, 3, 6 and 12 months after the sur gery. Results . The average duration of surgery was 8 minutes in the group without vitrectomy, 32 minutes in the group of subtotal vitrectomy and 18 minutes in the group of local vitrectomy. In the group of ERM removal without vitrectomy 6 patients out of 20 had a relapse of fibrosis found in follow-up period of 3 to 6 months. The groups showed a comparable improvement in visual acuity and photosensitivity of the central retina, as well as a decrease in the thickness of the r etina. In the group of subtotal vitrectomy, a significant IOP increase 1.6mmHg was recorded, as well as an increase in optic disc excavation, a more significant effect on the thickness of RNFL, indicators of the CFFF (drop from 38.4Hz to 34.3Hz), and a decrease in the light sensitivity of the retina periphery. In the local vitrectomy group, the IOP decreased significantly (from 16.5mmHg initially to 14.6mmHg). In addition, there was no decrease in the light sensitivity of the peripheral regions of the retina, and the changes in excavation of optic disc and RNFL were significantly smaller compared to the group of subtotal vitrectomy. Surgical cataract treatment was required in 47% of cases (14 patients) in the group of subtotal vitrectomy, and only 15% (5 patients) in the group of local vitrectomy. Patients of the first group did not require any cataract treatment. Conclusion . Subtotal vitrectomy with epiretinal membrane peeling is effective (30% increase in visual acuity), however, this intervention increases a risk of cataract development (47% of patients were operated on for cataract), negatively affects the optic nerve (CFFF decreased by 4Hz, increase of excavation of optic disk) and intraocular pressure (a 2mmHg persistent IOP increase 12 months later), and also, probably, reduces the sensitivity of the peripheral zone of the retina (-17dB in the upper-temporal quadrant). An alternative method of treatment is a removal of the membrane without vitrectomy. It allows to avoids the above-mentioned complications, but this method is associated with a high relapse rate (6 out of 20 patients had a relapse with a preservation complaints), due to the complexity of implementation and it significantly lim its the use of this method. Local vitrectomy demonstrated the most optimal balance between efficacy, safety, risk of recurrence and difficulty of implementation (no elevation of IOP, less pronounced negative impact on the optic disc and RNFL, no relapse, a low percentage of cataract development – 15%).