The reported incidence and prevalence of atrial fibrillation (AF) has been inconsistent across published studies. Using the National Health Insurance Service database of Korea, the prevalence and incidence of AF, and oral anticoagulation (OAC) use of AF patients were explored according to three different approaches; 'formal approach', considering individual AF diagnosis and mortality; 'limited diagnosis approach', using upper 5 main diagnosis; and 'medical use approach', using the number of medical use AF population by year without considering individual AF history and mortality. The AF prevalence progressively increased by 2.46-fold from 0.50% in 2004 to 1.54% in 2015 when using a 'formal approach' (p for trend <0.001). The overall prevalence was 1.09% and 0.97% when using a 'formal approach' and 'limited diagnosis approaches', respectively. Overall prevalence decreased to 0.52% with a 'medical use approach'. The trend of annual AF incidence was stable when using a 'formal approach', but increased by 15% when using a 'medical use approach'. OAC rate in 2015 was 2.1 times higher when using a 'medical use approach' compared to using a 'formal approach' (40.3% vs. 19.1%, p < 0.001). Given the wide variability in prevalence and incidence figures with different analysis approaches, careful attention to the analysis methodology is needed.