Today, clinical activity on the therapy and rehabilitation of patients with cognitive impairments is increasingly focused on cognitive rehabilitation based on cognitive training using digital technologies. These cognitive trainings are effective and have a combined mechanism of action: they contribute to how to improve the cognitive profile of patients, first of all, the elderly, and the long-term improvement of daily activity. The aim of the study – develop and implement a modern remote-based digital cognitive training method for geriatric patients with dementia-not-reaching dementia. Materials and methods. The study was conducted in outpatient conditions at the clinical base of the medical institute of Belgorod State University. The verification of cognitive deficits was carried out in the initial neuropsychological testing using a short scale of cognitive status research (COPS). The range of points for COPS is 27-24 (moderate cognitive impairment). Criteria for inclusion in the study: age 45-75 years, range of points for COPS - 27-24. During the study, two groups were formed: control and main. In each of them, by means of age randomization (average age 45-59 years, old age 60-74 years) are defined by 2 subgroups. The control group by random distribution included 50 elderly persons (total average age 66,7±0,9) and 50 middle-aged persons (total average age 52,2±0,5).The main group includes 50 elderly people (total average age 67,2±0,7) and 50 middle-aged people in (total average age 52,8±1,0). As a modern method of digital cognitive training based on remote application, a method of compiling an algorithm of action based on a personal computer was proposed for patients of the main group. The final protocol for testing the patient's cognitive domain included: a clock drawing test (THR) with a 10-point score scale and a COPS. The checkpoints were the 60th, 180th and 365th days. Results and their discussion. The analysis of the effectiveness of the algorithm of action was carried out on the 60th, 180th and 365th days of the study using COPS and TRH. According to the results of the study, middle-aged patients of the main group showed a reliable decrease in cognitive domain disorders on the 365th day (27,6±0,2) compared to the start of the program – 26,4±0,1 (p