To briefly review and discuss the literature on vitrectomy for diabetic macular oedema.Literature review.There is a copious literature on the subject of vitrectomy for diabetic macular oedema (DMO). The most commonly hypothesised mechanism for the potential benefit of vitrectomy is relief of vitreomacular traction; however, both transvitreal oxygenation and improved growth factor diffusion away from the premacular retina have also been suggested to be potentially beneficial effects. Other systemic and local factors including duration of oedema, extent of ischaemia and exudation, and extent of laser may result in permanent photoreceptor and capillary damage, which precludes anatomical or visual benefit. Much of the literature on the subject of vitrectomy for DMO is retrospective and uncontrolled but strongly suggestive of a benefit in terms of improved acuity and reduced macular thickness following vitrectomy. There are five published small randomised controlled trials on this subject. Taken as a whole, these studies do not suggest a benefit from surgery. Selection of patients for surgery on the basis of OCT partial vitreomacular separation or clinical signs of traction such as an epiretinal membrane or taut thickened hyaloid has been reported to be associated with a modest improvement in prospective studies but this has not been subjected to controlled study.The evidence at present suggests that vitrectomy for DMO should be restricted to those with clinical or OCT signs of traction.