The aim of this study was to assess the efficacy of cognitive training (CT), specifically computerised cognitive training (CCT) and virtual reality cognitive training (VRCT) programs for individuals living with mild cognitive impairment (MCI) or dementia and, therefore, at high risk of cognitive decline. A range of academic databases were searched (CINHAL, PSYCinfo and Web of Science) and the studies evaluated (n=16) were categorised as: (i) CCT (n= 10); (ii) VRCT (n=3) and (ii) multi-modal interventions (n=3). Effect sizes were calculated but a meta-analysis was not possible because of the large variability of study design and outcome measures adopted. The cognitive domains of attention, executive function and memory (visual and verbal) showed the most consistent improvements. There were positive effects on psychological outcomes (n=6): significant reductions on depressive symptoms (n=3) and anxiety (n=2) and improved perceived use of memory strategy (n=1). Assessments of activities of daily living demonstrated no significant improvements (n=8). Follow-up studies (n=5) demonstrated long term improvements in cognitive and psychological outcomes (n= 3) and in the intervention groups there was a plateau effect of cognitive functioning compared to the cognitive decline experienced by control groups (n= 2). CCT and VRCT were moderately effective in long term improvement of cognition for those at high risk of cognitive decline. Total intervention time did not mediate efficacy. Future research needs to improve study design by including larger samples, longitudinal designs and a greater range of outcome measures, including functional and quality of life measures to assess the wider effect of CT on individuals at high risk of cognitive decline.