PURPOSE. To evaluate the effectiveness of pars plana vitrectomy (PPV) with removal of the internal limiting membrane (ILM) in diabetic patients with macular edema unresponsive to grid laser photocoagulation. METHODS. In this randomized controlled study, 20 eyes of 10 patients with diabetic macular edema unresponsive to grid laser photocoagulation were evaluated. PPV with ILM removal was performed randomly in one eye each of 10 patients and taken as the study group; the untreated fellow eyes were taken as the control group. Main outcome measures were foveal thickness changes measured with optical coherence tomography and preoperative and postoperative visual acuity. Mann-Whitney U, Wilcoxon, and chi-square tests were used in statistical analysis. RESULTS. The mean age of the patients was 61.5±6 years (range 51 to 71). All patients were followed up for 12 months. In the study group, mean foveal thickness was 391.3±91.6 μm preoperatively and 225.5±49.4 μm postoperatively (p=0.009). In the control group, mean foveal thickness was 356.2±140 μm at baseline and 318.4±111.1 μm at 12-month follow-up (p=0.138). Mean decrease in foveal thickness was 165.8±114.8 μm in the study group and 37.8±71.2 μm in the control group (p=0.016). In the study group, best-corrected log-MAR visual acuity was 0.71±0.43 preoperatively and 0.54±0.45 postoperatively (p=0.125). In the control group, best-corrected logMAR visual acuity was 0.43±0.44 at baseline and 0.59±0.55 at 12-month follow-up (p=0.235). In the study group, visual acuity improved by two or more lines in 4 eyes (40%) and remained stable in 6 eyes (60%). In the control group, visual acuity improved by two or more lines in 1 eye (10%) and decreased by two or more lines in 3 eyes (30%). CONCLUSIONS. PPV with ILM removal appears to be an effective procedure for reducing diabetic macular edema unresponsive to grid laser photocoagulation. A further study with a large number of patients is required to assess the effectiveness and safety of this procedure. © Wichtig Editore, 2006.