The object of this article was to estimate the incidence rate of chronic atrial fibrillation (AF) in a general practice setting, to identify factors predisposing to its occurrence, and to describe treatment patterns in the year following the diagnosis. The method used was a population-based cohort study using the General Practice Research Database (GPRD) in the UK. We identified patients aged 40-89 years with a first ever recorded diagnosis of AF. The diagnosis was validated through a questionnaire sent to the general practitioners. A nested case-control analysis was performed to assess risk factors for AF using 1,035 confirmed incident cases of chronic AF and a random sample of 5,000 controls from the original source population. The incidence rate of chronic AF was 1.7 per 1,000 person-years, and increased markedly with age. The age adjusted rate ratio among males was 1.4 (95% CI 1.2-1.6). The major risk factors were age, high BMI, excessive alcohol consumption, and prior cardiovascular comorbidity, in particular, valvular heart disease and heart failure. Digoxin was used in close to 70% of the patients, and close to 15% did not receive any antiarrhythmic treatment. Close to 40% did not receive either warfarin or aspirin in the 3 months period after the diagnosis. Among the potential candidates for anticoagulation only 22% of those aged 70 years or older were prescribed warfarin in comparison to 49% among patients aged 40-69 years. Chronic AF is a disease of the elderly, with women presenting a lower incidence rate than men specially in young age. Age, weight, excessive alcohol consumption, and cardiovascular morbidity were the main independent risk factors for AF. Less than half of patients with chronic AF and no contraindications for anticoagulation received warfarin within the first trimester after the diagnosis.