Abstract: Objective: To explore the effects of cognitive training on type 2 Diabetes Mellitus (T2DM) with Mild Cognitive Impairment (MCI) in community. Methods: The T2DM patients with MCI from two communities in Changsha city were selected as study subjects from July, 2014 to January, 2015, they were randomly divided into experimental group (39 cases) and control group (40 cases). The cases in control group received routine care of the community, while the cases in experimental group got 3-month cognitive training combined with routine care of the community. Cognitive functions of the two groups were evaluated prior to intervention, after intervention and after 3-month follow-up using Montreal Cognitive Assessment (MoCA)Beijing version and Mini Mental State Examination (MMSE). Results: There were significant differences in the main effects of intervention, the main effects of time and interaction between intervention and time on scores of MoCA Beijing version and MMSE between two groups, respectively (P<0.05). There was no significant difference in scores of MoCA Beijing version and MMSE between experimental group and control group before intervention (P>0.05). There were significant differences in scores of MoCA Beijing version and MMSE between experimental group and control group after intervention and 3-month followup(P<0.05). In experimental group, the scores of MoCA Beijing version and MMSE after intervention and follow-up were significantly higher than those before intervention, respectively (P<0.05). The scores of visuospatial abilities, abstraction and delayed memory of MoCA in the experimental group were statistically higher than control group after intervention (P<0.05).Conclusion: Cognitive training could significantly improve cognitive functions of T2DM patients with MCI in community.