This review covers a seminal study of the relation between exposure to the glucagon-like peptide 1 (GLP-1) agonist semaglutide and incident non-arteritic anterior ischaemic optic neuropathy (NAION) in a neuro-ophthalmology clinic setting, subsequent studies in unselected populations, a meta-analysis of clinical trials and pathophysiology studies of the optic disc and retina that may help elucidate the relation between semaglutide and NAION in patients with diabetes or obesity. In the matched cohort study of neuro-ophthalmology patients, those treated using agents other than semaglutide had NAION rates, orders of magnitude higher than a background population, presumably because referral patterns led to the enrichment of the study populations with patients at high risk of NAION. With semaglutide, the rates of NAION were 4.28 and 7.64 times higher for type 2 diabetes (T2D) and obesity, respectively, and onset of NAION was within 14 months of treatment initiation, whereas non-semaglutide NAION was evenly distributed over the 3 years of follow-up. Of four health care registry studies, each covering more than 100 000 patients, two found relative rates of NAION two to three times higher with semaglutide than without semaglutide, one found a trend towards semaglutide being associated with NAION in patients with type 2 diabetes only, and one found statistically insignificant imbalances between the two alternatives. The meta-analysis of various GLP-1 receptor agonists versus placebo or active comparator found no significant difference in rates of NAION. Prior reports of long-term glycaemia reduction being associated with early worsening of retinopathy and with NAION indicate that semaglutide may promote such events in proportion to its antihyperglycaemic potency. The association of NAION with small, crowded discs, optic disc oedema and peripapillary exudation suggests that semaglutide-related NAION may result from changes in perfusion that lead to venous dilation and, presumably, to venous congestion during relative hypoglycaemia. Given the retrospective nature of the epidemiological studies, causality cannot be inferred, but a cautious approach to the use of semaglutide and other powerful glycaemia-reducing agents seems warranted, particularly in patients with crowded optic discs, a characteristic that can be identified by proactive eye examination for disc-at-risk characteristics by the use of optical coherence tomography.