The main cause of disability in patients with diabetes mellitus (DM) is its micro- and macrovascular complications. Diabetic retinopathy (DR) is a microvascular complication of DM and one of the leading causes of blindness and low vision in the world among people of working age. The aim of this study was to review the domestic and foreign literature reflecting the significance of early digital screening for diabetic retinopathy, as well as the experience of using screening programmes in the world and the Russian Federation. Materials and methods: the data of domestic and foreign scientific literature were used. Research methods: monographic, analytical. Scientific articles were searched in bibliographic databases PubMed (Medline), in the search database of scientific publications eLIBRARY and GOOGLE search engine. Results. According to the literature, only 33-68% of adults diagnosed with diabetes mellitus undergo annual retinal screening. Barriers to screening at the appropriate time include low health literacy, i.e. patients do not associate their underlying disease with the risk of developing blindness, not understanding the importance of eye examination, economic issues, lack of time, lack of transport, reluctance to schedule follow-up appointments, and systemic factors in the organisation of care. AI-assisted retinal image analysis can further simplify screening, reducing the burden on medical staff. Neural networks are highly efficient, allowing computer systems to analyse data, greatly facilitating the work of the clinician.A new period in the diagnosis of DR using AI systems began in 2013. The published results of the first pre-registered clinical trial of an AI system conducted by Abramoff et al in 2013 eventually led to the US Food and Drug Administration (FDA) authorising the marketing of a medical device that uses AI to detect DR in adults with DM (IDx-DR) as soon as possible.Thus, in 2018, the IDx-DR fully autonomous diagnostic system was the first approved autonomous AI programme. The Russian Federation has all the possibilities for screening: detection of biological markers of retinal lesions, photoregistration of the ocular fundus, including the use of telemedicine technologies, artificial intelligence technologies, fluorescence angiography and optical coherence tomography, including OCT in angioregime, but currently there is no national programme for DR screening and no clear system of organisation. Conclusion. The review of domestic and foreign literature sources reflecting the significance of early digital screening in DR, as well as the experience of using screening programmes in the world and in the Russian Federation convincingly proves the effectiveness and expediency of early diagnosis of DR and the use of disease screening. The World Health Organisation stresses the importance of DR screening