Objectives. Digital health interventions enable services to support people living with dementia and Mild Cognitive Impairment (MCI) remotely. This literature review gathers evidence on the effectiveness of digital health interventions on physical, cognitive, behavioural and psychological outcomes, and Activities of Daily Living in people living with dementia and MCI. Methods/Design. Searches, using nine databases, were run in November 2021. Two authors carried out study selection/appraisal using the Critical Appraisal Skills Programme checklist. Study characteristics were extracted through the Cochrane handbook for systematic reviews of interventions data extraction form. Data on digital health interventions were extracted through the template for intervention description and replication (TIDieR) checklist and guide. Intervention effectiveness was determined through effect sizes. Meta-analyses were performed to pool data on intervention effectiveness. Results. Twenty studies were included in the review, with a diverse range of interventions, modes of delivery, activities, duration, length, frequency, and intensity. Compared to controls, the interventions produced a moderate effect on cognitive abilities (SMD = 0.36; 95% CI = -0.03 – 0.76; I2 = 61%), and a negative moderate effect on basic ADLs (SMD = -0.40; 95% CI = -0.86 – 0.05; I2 = 69%). Stepping exergames generated the largest effect sizes on physical and cognitive abilities. Supervised training produced larger effect sizes than unsupervised interventions. Conclusion. Supervised intervention delivery is linked to greatest benefits. A mix of remote and face-to-face delivery could maximise benefits and optimise costs. Accessibility, acceptability and sustainability of digital interventions for end-users must be pre-requisites for the development of future successful services.