Background: Diabetic macular edema (DME) is a leading cause of visual impairment today, and occurs in approximately 10% of diabetic patients and 29% of those with disease duration of more than 20 years. Objective: To detect the effectiveness of Pars Plana Vitrectomy PPV in treatment of refractory DME, and the differences in the anatomical and functional outcomes between PPV with and without internal limiting membrane (ILM) peeling. Patients and methods: This was a prospective comparative study of 50 eyes of patients with refractory DME divided into 2 equal groups: Group A underwent vitrectomy without ILM peeling, and Group B underwent vitrectomy with ILM peeling.      All patients were subjected to full history and ophthalmic examination including Uncorrected / best corrected visual acuity (UCVA/BCVA) expressed in Decimal units, refraction using automated refractometer (Topcon KR-800 Auto refractometer), intraocular pressure (IOP) measurement by Goldman Applanation tonometer, Slit lamp biomicroscopy to assess cornea, depth of anterior chamber, state of pupil dilatation, lens morphology, dilated pupil fundus examination, Fundus Fluorescein Angiography (FFA) to show Features of diabetic maculopathy, Optical Coherence Tomography (OCT) To document macular thickness  and Multifocal Eletroretinogram (MF ERG) to assess the electrical response in central retinal area. The study was done at Al-Azhar University Hospitals between May 2019 and February 2021. Results: The mean BCVA in our patients improved from 0.16±0.07 pre-operatively to 0.29±0.11 post-operatively in group A, and from 0.20±0.10 pre-operatively to 0.36±0.14 post-operatively in group B. This showed statistically significant improvements in both groups. The mean CMT in our patients improved from 495.64±113.37 μm m pre-operatively to 323.24± 63.21 μm post-operatively in group A and from 515.20±82.47 μm m pre -operatively to 292.96±59.33 μm m post-operatively in group B, and showed statistically significant improvements in both groups. P1 amplitude in ring (1+2) of the MF-ERG improved from 33.36±9.65 nv/deg2 preoperatively to 43.00±9.51 nv/deg2 at the end of the 3rd month in group A, In Group B, it improved from 35.40±10.99 nv/deg2 preoperatively to 42.16±11.84 nv/deg2 at the end of the 3rd month. Compared to preoperative values, there were statistically significant differences in both groups.  P1 latency in ring (1+2) of the MF-ERG improved from 51.12±11.43ms preoperatively to 36.12±8.28 ms at the end of the 3rd month, in group A. In Group B, it improved from 54.85±12.03 ms preoperatively to 39.60±10.09 ms at the end of the 3rd month. Compared to preoperative values there were statistically significant differences in both groups. Conclusion: PPV with and without ILM peeling improved refractory DME even without evidence of VMT. This improvement was: structural (anatomical) improvement, i.e decrease in the CMT, functional improvement, i.e improvement in the BCVA, and macular response detected by MF ERG.