Background Atrial fibrillation (AF) is a common arrhythmia associated with increased cardiovascular morbidity and mortality. This study was undertaken to analyze the epidemiological factors and evaluate the current status of treatment in patients with AF in Mainland China. Methods Retrospective analysis of hospital records were taken from patients with primary diagnosis of AF, discharged from January 1999 to December 2001. A total of 9297 cases (mean age 65.5 years) with AF were enrolled from 40 hospitals in major parts of China. Results The percentage of hospital admissions with AF was gradually increased comparing to those of total cardiovascular admission during three years, with the average of 7.9%. The cases distribution progressively rose with age. The causes and associated conditions of AF: advanced age 58.1%, hypertension 40.3%, coronary heart disease 34.8%, heart failure 33.1%, rheumatic valvular disease 23.9%, idiopathic AF 7.4%, cardiomyopathy 5.4%, diabetes 4.1%. The most common coexistence among these variables was advanced age with hypertension. Permanent AF almost accounted for half of these cases (49.5%), paroxysmal and persistent AF were 33.7% and 16.7%, respectively. Paroxysmal AF was mainly treated with rhythm control (56.4%). However, 82.8% of patients with chronic AF had therapeutic strategy of rate control. In patients with persistent AF, the cardioversion had been attempted in cases more than 50%, with only 31.1% of these patients who could maintain stabilized sinus rhythm. The prevalence of stroke in this group was 17.5%. In nonvalvular AF patients the risk factors that significantly associated with stroke included advanced age, history of hypertension, coronary heart disease and type of AF. Sixty-four point five percent of these patients received antithrombotic therapy with dominated use of antiplatelet agents. The long-term prevention with anticoagulants only accounted for 6.6%. In this investigation patients with antiplatelets as well as patients with anticoagulants showed significant lower stroke rate in comparison with those managed neither. However, the difference between antiplatelets and anticoagulants in terms of stroke rate was not significant. Conclusions Most epidemiological factors of AF from this group showed highly in accordance with those from the reports from other countries, such as age distribution, causes and associated conditions, type of AF, dominantly with approach of rate control. Both antiplatelet and anticoagulant treatments significantly reduced stroke rate. But there was no significant difference between these two kinds of treatments in reducing stroke rate.