Purpose: To report anatomic and visual outcomes of vitrectomy and indocyanine green (ICG)-assisted peeling of the retinal internal limiting membrane (ILM) in the treatment of diffuse diabetic macular edema. Methods: In a retrospective in terventional case series, 15 eyes of 11 patients with refractory diffuse diabeti c macular edema underwent pars plana vitrectomy with removal of the ILM, which w as stained by intravitreal injection of ICG (0.1-0.2 ml of 0.5%ICG), performed by a single surgeon. The patients were followed up for 14-28 months (mean 20.5 months). The main outcome measures were assessment of macular edema by optical coherence tomography and determination of visual acuity and visual field. Result s: Intravitreal ICG visualized the ILM to facilitate complete removal of the str ucture. Qualitative assessment of optical coherence tomography images at the end of follow-up revealed that retinal thickness in the macula appeared nearly nor mal with or without reappearance of foveal pit in 11 of the 15 eyes (73.3%), de creased in 3 eyes (20.0%), and did not change in 1 eye (6.6%). Best-corrected visual acuity at the end of follow-up improved by 2 lines or more in 4 eyes (2 6.7%), virtually unchanged in 6 eyes (40.0%), and deteriorated by 2 lines or m ore in 5 eyes (26.7%). The mean logMAR visual acuity was 0.680 (approximately 1 2/60) preoperatively and 0.812 (approximately 9/60) postoperatively, the differe nce being not statistically significant (paired t-test, P=0.445). Seven (46.7% ) of the 15 eyes developed optic nerve atrophy that occurred gradually within 6 months after surgery and caused irreversible peripheral visual field defect pred ominantly affecting the nasal field. Conclusion: Intravitreal application of ICG is beneficial in uneventful ILM peeling to help resolution of diffuse diabetic macular edema, but it may potentially damage the optic nerve fibers and lead to unfavorable visual outcomes. Purpose: To report anatomic and visual outcomes of vitrectomy and indocyanine green (ICG) -assisted peeling of the retinal internal limiting membrane (ILM) in the treatment of diffuse diabetic macular edema. Methods: In a retrospective in terventional case series, 15 eyes of 11 patients with refractory diffuse diabeti c macular edema underwent pars plana vitrectomy with removal of the ILM, which w as stained by intravitreal injection of ICG (0.1-0.2 ml of 0.5% ICG), performed by a single surgeon. The patients were followed up for 14-28 months (mean 20.5 months). The main outcome measures were assessment of macular edema by optical coherence tomography and determination of visual acuity and visual field. Results s: Intravitreal ICG visualized the ILM to facilitate complete removal of the strcture. Qualitative assessment of optical coherence tomography images at the end of follow-up revealed that retinal thickness in the macula appeared nearly nor mal with or without reappearance of fovea Best-corrected visual acuity at the end of follow-up improved by 2 (73.3%), de creased in 3 eyes (20.0%), and did not change in 1 eye lines or more in 4 eyes (2 6.7%), virtually unchanged in 6 eyes (40.0%), and deteriorated by 2 lines or m ore in 5 eyes (26.7%). The mean logMAR visual acuity was 0.680 (approximately 1 2 / Seven (46.7%) of the 15 eyes developed optic nerve atrophy that had gradable within 6 months (60%) preoperatively and 0.812 (approximately 9/60) postoperatively, the difference not quite unique significant (paired t- after surgery and caused irreversible peripheral visual field defect predominantly affecting the nasal field. Conclusion: Intravitreal application of ICG is beneficial in uneventful ILM peeling to help resolution of diffuse diabetic macular edema, but it may may damage the optic nerve fibers and lead to unfavorable visual outcomes.