Optical coherence tomography (OCT) of the retina and around the optic nerve head and corneal confocal microscopy (CCM) are non-invasive and repeatable techniques that can quantify ocular neurodegenerative changes in individuals with diabetes. We systematically reviewed studies of ocular neurodegenerative changes in adults with type 1 or type 2 diabetes and noted changes in the retina, the optic nerve head, and the cornea. Of the 30 studies that met our inclusion criteria, 14 used OCT and 16 used CCM to assess ocular neurodegenerative changes. Even in the absence of diabetic retinopathy, several layers in the retina and the mean retinal nerve fibre layer around the optic nerve head were significantly thinner (-5·36 μm [95% CI -7·13 to -3·58]) in individuals with type 2 diabetes compared with individuals without diabetes. In individuals with type 1 diabetes without retinopathy none of the intraretinal layer thicknesses were significantly reduced compared with individuals without diabetes. In the absence of diabetic polyneuropathy, individuals with type 2 diabetes had a lower nerve density (nerve branch density: -1·10/mm(2) [95% CI -4·22 to 2·02]), nerve fibre density: -5·80/mm(2) [-8·06 to -3·54], and nerve fibre length: -4·00 mm/mm(2) [-5·93 to -2·07]) in the subbasal nerve plexus of the cornea than individuals without diabetes. Individuals with type 1 diabetes without polyneuropathy also had a lower nerve density (nerve branch density: -7·74/mm(2) [95% CI -14·13 to -1·34], nerve fibre density: -2·68/mm(2) [-5·56 to 0·20]), and nerve fibre length: -2·58 mm/mm(2) [-3·94 to -1·21]). Ocular neurodegenerative changes are more evident when diabetic retinopathy or polyneuropathy is present. OCT and CCM are potentially useful, in addition to conventional clinical methods, to assess diabetic neurodegenerative changes. Additional research is needed to determine their incremental benefit and to standardise procedures before the application of OCT and CCM in daily practice.