Introduction: Evidence using real-world data is sparse regarding the effects of oral anticoagulants (OACs) among patients with kidney disease. The aim of this study was to investigate the effects of kidney disease on ischemic stroke (IS) or systemic embolism (SE) among patients taking OAC, using large-scale real-world data in Japan. Methods: This was a retrospective cohort study using claims data and health checkup data from health insurance associations in Japan, from January 2005 to June 2017. We enrolled 21,581 patients diagnosed with atrial fibrillation (AF). Of the total population, 11,848 (54.9%) patients were taking OAC. A Cox proportional hazards model was used to examine the effect of kidney disease on IS/ SE with or without OAC. Results: During follow-up, 208 participants who were not taking OAC (mean follow-up 2.6 years) and 200 who were taking OAC (mean follow-up 3.0 years) experienced IS/SE. The % IS/SE incidence rates with and without kidney disease were 2.42/person-year and 0.63/person-year in the total population, 3.66/person-year and 0.76/personyear in the group without OAC use, and 1.52/person-year and 0.55/person-year in patients with OAC use, respectively. Hazard ratios (HRs) and 95% confidence intervals (CIs) of kidney disease for IS/SE were high, irrespective of OAC, even after adjustment: adjusted HR 2.62 (95% CI: 1.72-3.99) without OAC and adjusted HR 2.03 (95% CI: 1.20-3.44) with OAC; p = 0.193 for interaction between no OAC and OAC. Although bleeding risk was also high for kidney disease irrespective of OAC use (HR 2.93 [95% CI: 2.27-3.77] in the total population, HR 3.08 [95% CI: 2.15-4.43] in the group without OAC, and HR 2.73 [95% CI: 1.90-3.91] in the group with OAC use), net clinical benefit indicated that the benefit of OAC use exceeded the risk of bleeding: HR 4.50 (95% CI: 0.76-8.23) among those with kidney disease and HR 0.35 (95% CI: 0.04-0.66) among those without kidney disease. Conclusion: Although we found that OAC use was effective and recommended for patients with AF, advanced kidney disease is still an independent risk factor for IS/SE, even in patients taking OAC. Physicians should be aware of this risk and strictly control modifiable risk factors, regardless of OAC use.