Objective To investigate the characteristics of mild cognitive impairment (MCI) in the elderly patients with non-valvular atrial fibrillation and its affecting factors. Methods We retrospectively analyzed the clinical data of 104 elderly inpatients in the Department of General Medicine of Tongji Hospital of Huazhong University of Science and Technology from October 2018 to September 2019, with 52 cases in atrial fibrillation group (AF group) and non-atrial fibrillation group (non-AF group)respectively. Cognitive function was assessed using mini-mental state evaluation (MMSE) and Montreal cognitive assessment (MoCA) for the detection rate, clinical characteristics and risk factors of MCI in both groups. SPSS statistics 26.0 was used for data analysis. According to the data type, t-test, non-parametric rank sum test or χ² test was used for comparison between groups. Logistic regression model was used for multivariate analysis. Results Compared with the non-AF group, the AF group had lower total MMSE [(27.37±1.53) vs (28.33±1.01)] and MoCA [(24.71±1.54) vs (26.29±1.63)]; and significantly higher detection rate of MCI (MMSE:38.5% vs 7.7%; MoCA:61.5% vs 30.8%), the difference being statistically significant (P<0.05). Evaluation of 7 sub-cognitive domains using the MoCA showed that the orientation ability, visual space and executive function, language ability, and abstract generalization in the AF group were worse than in the Non-AF group, the difference being statistically significant (P＜0.05). MCI patients were defined as MoCA being positive, and five risk factors (atrial fibrillation, hypertension, hyperlipidemia, age, education level) that had an impact on MCI both by the MMSE and MoCA were selected as independent variables for logistic regression analysis. The results showed that longer schooling years were protective factors for MCI, while atrial fibrillation, advanced age and hypertension were independent risk factors for MCI, of which atrial fibrillation was a stronger risk factor (OR=3.163,95%CI 1.190-8.406; P＜0.05). Conclusion Detection rate of MCI increases significantly in the elderly patients with non-valvular atrial fibrillation, and their cognitive impairment mainly affects orientation, visual space and executive function, language ability and abstract generalization. Atrial fibrillation, age and hypertension have an impact on MCI, and effective management of risk factors such as atrial fibrillation can play a positive role in the prevention and treatment of MCI.