Background/Aims: To investigate the efficacy of internal limiting membrane (ILM) peeling following pars plana vitrectomy for cases of recalcitrant diabetic macular edema in terms of improved visual acuity and reduced central macular thickness (CMT). Materials and Methods: A prospective randomized interventional study of 60 eyes with recalcitrant diabetic macular edema between age groups of 30–75 years of either sex were included if they are unresponsive to at least one intravitreal steroid after 4 weeks followed by at least one intravitreal antivascul©ar endothelial growth factors followed by conventional laser treatment after 8 weeks at least within or <8 weeks before enrollment into the study. The absence of macular ischemia and CMT of more than 350 μ were included. The main outcome measures were corrected and uncorrected visual acuity, reduction of CMT, early and postoperative complications, and intraocular pressure. Results: The mean (standard deviation [SD]) of coherence tomography (OCT) was 558.3 (74.2) and 537.6 (52.6) in ILM and non-ILM group, respectively. On day 1, OCT was reduced to 419.0 in ILM peeling group (558.3–419.0, P< 0.001, statistically significant) and increased to 543.8 in non-ILM peeling group (537.6–543.8, P = 0.002, statistically significant) from baseline. The mean (SD) of best-corrected visual acuity (BCVA) at baseline in ILM peeling and non-ILM peeling were 1.24 (0.15) and 1.27 (0.15), respectively. The reduction in mean of BCVA at 6 and 12 weeks from baseline (1.24–1.16, P = 0.010 and 1.24–1.15, P = 0.014) was found to be statistically significant in ILM peeling whereas there is no change in the mean BCVA over 12 weeks. Conclusion: ILM peeling gives satisfactory results in recalcitrant diabetic macular edema.