AIMS/BACKGROUND: This study was performed to identify factors predictive for recurrence of idiopathic and secondary epiretinal membranes after vitrectomy. Long term visual outcome was determined and compared with eyes without recurrence of epiretinal tissue. METHODS: Out of 350 consecutive cases vitrectomised for removal of epiretinal macular membranes, 42 patients (42 eyes) with recurrence of epiretinal membranes were re-examined and their records reviewed. Recurrence of epiretinal membranes was determined by new appearance of epiretinal tissue or abnormal sheen of the central retina. The preoperative findings, best postoperative and final visual acuity of these patients were compared with a comparable cohort of eyes without recurrence after removal of macular pucker (131 eyes). The eligibility criterion was a follow up of at least 6 months. RESULTS: The best postoperative (0.46 (SD 0.24)) and the final visual acuity (0.39 (0.26)) of eyes with recurrent membranes (42 eyes) was statistically better than preoperative vision (0.27 (0.22)) (p = 0.0003 and 0.0089). Because of reduced or distorted vision eight of the 42 eyes underwent a revitrectomy during the follow up (mean 23 months). Best postoperative visual acuity (0.50 (0.21)) and final visual acuity (0.47 (0.25)) of eyes which received a second vitrectomy (n = 8) were not statistically different from the 32 eyes with recurrence of epiretinal tissue but without revitrectomy (p = 0.253 and 0.343). In addition, eyes with recurrence showed no different visual outcome compared with eyes without recurrence of epiretinal membranes after vitrectomy (p = 0.84). Recurrence of epiretinal membranes was correlated with neither the type of membrane (thick or thin) nor incomplete membrane removal during pars plana vitrectomy. CONCLUSION: This study demonstrates the overall favourable visual prognosis of patients with vitrectomy for macular pucker, even with recurrence of the premacular membrane.