Aims: To assess the effect of oral anticoagulant (OAC) administration on incidence of dementia in patients with atrial fibrillation (AF) with Systematic review and meta-analysis in according with the Preferred Reporting Items for Systematic Review and Meta-analysis Protocols. Methods: We systematically searched the electronic databases including Pubmed, Embase, Cochrane library, and ClinicalTrails.gov for relevant articles. The primary outcome was the incidence of dementia. The adjusted risk ratio (RR), odds ratio, or hazard ratio were extracted and pooled by the random-effects models. Subgroup analysis was performed according to the setting observational window. Risk of bias was assessed using the Newcastle-Ottawa Scale, while publication bias was assessed by the Begg’s and Egger’s tests. Results: Nine studies included in this review (2 prospective and 7 retrospective observational studies, including 613,920 patients). The results presented the significant association between OAC therapy and the reduced risk of dementia compared with no treatment (RR [95%CI] =0.72 [0.60, 0.86], I2=97.2%; P =0.000). In the subgroup analysis, the pooled RR became statistically non-significant (including four studies, RR [95%CI] =0.75 [0.51, 1.10], I2=98.8%; P =0.000). There is no significant risk of bias and publication bias. Conclusions: This study indicated the protective effect of OAC therapy for dementia in patients with AF. However, the results are limited because of high heterogeneity, inconsistent direction of effect in subgroup analysis. Further prospective well-designed study is needed with longer follow-up duration in younger patients.