Purpose: To evaluate the effect of internal limiting membrane peeling and air tamponade for idiopathic macular hole, and explore reasons and interventions for persistent holes. Methods: One hundred and thirty-five eyes with Stage III and IV idiopathic macular hole that underwent 23-gauge vitrectomy, internal limiting membrane peeling, and air tamponade were reviewed. Eyes with persistent holes underwent a second surgery. Outcome-related factors and interventions treating persistent holes were discussed. Results: The initial closure (Type I) rate was 89.63% (121/135). Eyes that underwent the second surgery all obtained final closure (Type I). Diameter of macular hole was significantly smaller (P < 0.001) and duration of symptoms was significantly shorter (P = 0.017) in initially closed cases than in unclosed ones. Binary logistic regression indicated large diameter of macular hole as a risk factor for initial closure (P = 0.004). A cutoff value of 677 mu m was provided by receiver operating characteristic curve to predict initial closure (P < 0.001). Best-corrected visual acuity of all individuals improved significantly (P < 0.001) from 20/154 to 20/40 (mean follow-up: 4.5 months). Conclusion: Internal limiting membrane peeling and air tamponade for idiopathic macular hole provide satisfactory morphologic and functional outcomes. Large diameter of macular hole and long duration of symptoms are risk factors for initial closure. Proper second surgery can obtain satisfactory outcomes for persistent holes.