The authors report on the three-year outcomes of the Ozurdex Dexamethasone (DEX) intravitreal implant multi-centre trial for the treatment of diabetic macular oedema (DME). Patients with DME (best-corrected vision of 20/50 to 20/200 and central retinal thickness of ≥300µm) were randomised in a 1:1:1 ratio to DEX implant 0.7mg, DEX implant 0.35mg or sham. Exclusion criteria included uncontrolled diabetes, concurrent use of systemic steroid, glaucoma, active neovascularisation, history of intraocular laser or surgery within 90 days, intravitreal anti-VEGF injection within 90 days, and intravitreal triamcinolone injection within 180 days. There were 1048 patients enrolled, but only 607 (57.9%) completed the three year study duration. The average number of treatments was 5.0 in the DEX implant 0.7mg group, 5.2 in the DEX implant 0.35mg group, and 5.1 in the sham group. In terms of visual acuity improvement of >15 letters, this was achieved in 22.2% of the DEX implant 0.7mg group, in 18.4% of the DEX implant 0.35 group, and in 12.0% of the sham group. The average reduction in central retinal thickness was 111.6µm in the DEX 0.7mg group, 107.9µm in the DEX 0.35mg group, and 41.9µm in the sham group. The commonest complications were elevated intraocular pressure (IOP) and cataract. IOP elevation requiring topical medication occurred in 41.5% in the DEX 0.7mg group, 37.6% in the DEX 0.35mg group, and 9.1% in the sham group. Cataract surgery was required in 59.2% of phakic eyes. In comparison to sham, the Ozurdex intravitreal implant show significant visual improvement and reduction in central retinal thickness. The side-effect profile was less in comparison to those previously reported for intravitreal fluocinolone and triamcinolone, but still significant nonetheless. A trial comparing Ozurdex to intravitreal anti-VEGF injections for DME will help to establish whether the benefit of reduced numbers of injections is outweighed by the significant side-effect profile.