Background and Purpose: Atrial fibrillation is a risk factor for dementia, and oral anticoagulant use is associated with a decreased risk of dementia in patients with atrial fibrillation. We aimed to investigate whether the risk of dementia would be different between patients treated with direct oral anticoagulants (DOACs) compared with those with warfarin. Methods: Using the Korean nationwide claims database from January 2014 to December 2017, we identified oral anticoagulant–naive nonvalvular atrial fibrillation patients aged ≥40 years. For the comparisons, warfarin and DOAC groups were balanced using the inverse probability of treatment weighting method. The primary outcome was incident dementia. Results: Among 72 846 of total study patients, 25 948 were treated with warfarin, and 46 898 were treated with DOAC (17 193 with rivaroxaban, 9882 with dabigatran, 11 992 with apixaban, and 7831 with edoxaban). During mean 1.3±1.1 years of follow-up, crude incidence of dementia was 4.87 per 100 person-years (1.20 per 100 person-years for vascular dementia and 3.30 per 100 person-years for Alzheimer dementia). Compared with warfarin, DOAC showed a comparable risks of dementia, vascular dementia, and Alzheimer dementia. In subgroup analyses, DOAC was associated with a lower risk of dementia than warfarin, particularly in patients aged 65 to 74 years (hazard ratio, 0.815 [95% CI, 0.709–0.936]) and in patients with prior stroke (hazard ratio, 0.891 [95% CI, 0.820–0.968]). When comparing individual DOACs with warfarin, edoxaban was associated with a lower risk of dementia (hazard ratio, 0.830 [95% CI, 0.740–0.931]). Conclusions: In this large Asian population with atrial fibrillation, DOAC showed a comparable risk of dementia with warfarin overall. DOACs appeared more beneficial than warfarin, in those aged 65 to 74 years or with a history of stroke. For specific DOACs, only edoxaban was associated with a lower risk of dementia than warfarin.