It's widely acknowledged that, as a neurodegenerative aging disease representing an intermediate stage between cognitive intactness and Alzheimer's disease (AD), Mild cognitive impairment (MCI) poses an excessive burden on patients' well-being, family members, health-care providers as well as the whole society. This study focuses on three cognitive interventions proposed by Clare and Woods, which are, Cognitive stimulation (CS), Cognitive training (CT) and Cognitive rehabilitation (CR). Our Network meta-analysis (NMA) aims to compar them with one another to determine the optimal cognitive intervention for elderly adults with MCI in improving their cognitive function. We applied extensive strategies to preliminary literature retrieval to identify relevant randomized controlled trials (RCTs) which scrupulously compared any two of the three cognitive interventions with one another or any one of the three with a control group as the placebo or non-active group in treating elder patients with MCI in accordance with Petersen's criteria. Our NMA of cognitive interventions for patients diagnosed with MCI appraised the relative effectiveness of cognitive interventions across trials simultaneously. Our study attempts to summarize available data to suggest that CS (Mean difference [MD] = 0.95, 95% confidence interval [CI]:0.27, 1.70) and CT (MD = 0.70, [CI]:0.11,1.30) were significantly beneficial to MCI patients for improving their cognition status while CR (MD = 0.59, [CI]:-0.30,1.50) scored lowest. Our study suggested CS was most likely to be the best intervention for improving the cognitive function of MCI patients.