Background: To analyze the impact of using different renal function equations and stroke prevention strategy in atrial fibrillation (AF) across all chronic kidney disease (CKD) stages. Methods: We used the Cockcroft-Gault (CG), Modified Diet in Renal Disease (MDRD), and Chronic Kidney Disease Epidemiology Collaboration (CKD-EPI) equations to classify 39,217 patients into stage 1 to 5 CKD during July 1st, 2001, and September 30st, 2018. The endpoint is a composite outcome including ischemic stroke or major bleeding or mortality. Results: More patients belonged to stage 1 and 2 CKD using the MDRD and CKD-EPI equations. In subgroups of patients with eGFR-MDRD or eGFR-CKD-EPI ? 60 mL/min, a 17-18% increase of event was observed in patients with eGFR-CG < 60 mL/min compared to those ? 60 mL/min. Compared to no oral anticoagulant (OAC), OAC use was associated with a significantly lower risk of event across stage 1 to 4 CKD but not in stage 5 CKD. Both warfarin and NOACs exhibited better outcome compared to no OAC across stage 1 to 4 CKD while NOACs was associated with more risk reduction compared to warfarin. Among patients on OACs, there was a trend toward better outcome with NOAC than warfarin across stage 2-4 CKD but not in stage 1 and 5 CKD. Conclusions: OAC should be used in stage 1 to 4 CKD with NOAC exhibiting the trend of better outcome through stage 2 to 4 CKD than warfarin. For stage 5 CKD, optimal strategy remains undetermined.