Many large-scale randomized control trials (RCTs) have been performed regarding treatment strategy in atrial fibrillation (AF) in Western countries and also in Japan. However, limited data are available concerning real-world relationships between the treatment strategy and prognosis of AF patients. Out of a prospective cohort of The Shinken Database 2004 (n = 2 412), 286 AF patients (male 205, 64.1 +/- 12.3 years, paroxysmal form 165) were retrospectively investigated. The percentage of AF patients under the rhythm control strategy was evaluated using the Kaplan-Meier method, which showed the cumulative proportion of rhythm control strategy was approximately 30% at the 90th day after the initial visit and 40.0% at 1 year. The average time to the first rhythm control strategy was 68.3 +/- 106.7 days. Those under rhythm control strategy were associated with fewer coexisting organic cardiac diseases, a younger age, and smaller left atrial dimension. Consequently, they showed very good prognosis (cumulative incidence rate of cardiovascular events at 1 year was 0.0%). Careful induction of rhythm control strategy, which was adopted in approximately 40% of the patients in the real world, was associated with fewer comorbidities and therefore might lead to better prognosis, although this does not mean the direct effects of rhythm control strategy.