Objective: Whilst exercise, cognitive, and dual-task interventions have been shown to improve cognitive function within a healthy aging population, it remains unclear as to what effect such interventions may have in a type 2 diabetes mellitus (T2DM) population. The primary aim of this research was to systematically review and quantify the effect of exercise, cognitive, and dual-task interventions for improving cognitive function in T2DM. Design: Systematic review/meta-analyses Methods: Databases (PubMed, EMBASE, CINAHL, Web of Science, ClinicalTrial.gov, Cochrane register of controlled trials, Prospero, HTA, and DARE) of published, unpublished, and ongoing studies were searched for randomised controlled trials investigating the effect of exercise, cognitive, and dual-task interventions upon cognitive function in T2DM. Results: This review identified three studies investigating the effects of an exercise intervention and one study investigating the effect of a cognitive intervention upon cognitive function in T2DM. Meta-analyses indicated a significant effect of exercise for improving global cognitive function (mini-mental state examination SMD=0.22, 95% confidence interval 0.03 – 0.41, P=0.03) and inhibitory control (Stroop task SMD=0.54, 95% confidence interval -1.00 – -0.08, P=0.02) but not working memory (digit symbol SMD=0.09, 95% confidence intervals -0.10 – 0.28, P=0.35). Calculated effect sizes of outcome measures in the cognitive study indicated a beneficial effect of cognitive training upon cognitive function in T2DM. The risk of bias assessment in this review was hindered predominantly by poor reporting practices of included studies. Due to incomplete reporting of methodological procedures, two studies were judged to have a high risk of overall bias whilst the remaining two were judged as having a moderate overall risk of bias. Conclusion: The findings of the present systematic review and meta-analyses provide preliminary evidence for exercise and cognitive interventions improving cognitive function in T2DM. The poor reporting practices of included studies means that future research in this area should identify relevant reporting guidelines (e.g. CONSORT) to reduce the risk of bias and facilitate transparent reporting.