The purpose of this investigation was to define cardioversion success rates, frequency of complications of cardioversion, and current treatment practices in elderly patients (aged > or = 65 years) with atrial fibrillation (AF). The results were compared with those in younger patients (aged < 65 years). The investigation was a prospective multicenter observational study with 61 participating cardiology clinics. Consecutive patients in whom cardioversion of AF was planned had to be prospectively registered. Of 1,152 patients registered, 570 (49.5%) were < 65 years old (group 1) and 582 (50.5%) were > or = 65 years (group 2). The overall success rate of cardioversion on an intention-to-treat basis was 76.1% in group 1 and 72.7% in group 2 (p = 0.18). In multivariate analysis, left atrial size and New York Heart Association functional class before cardioversion were identified as predictors of success (p < 0.001, respectively; p = 0.025). These clinical factors were not equally distributed between the age groups: Left atrial size was larger in the elderly than in younger patients (44.0 +/- 6.4 mm vs 42.8 +/- 6.4 mm; p = 0.006) and a New York Heart Association functional class > or = II was more prevalent in group 2 than in group 1 (48.6% vs 29.6%; p < 0.001). The overall complication rates were not significantly different between the 2 groups (4.2% in group 1 vs 5.3% in group 2; p = 0.37). The frequency of patients who were adequately anticoagulated for cardioversion was 56.9% in age group 1 and 39.6% in age group 2 (p < 0.001). In chronic AF the same trend for age-dependent underuse of anticoagulation was observed. Age itself was not a predictor of cardioversion success and did not predispose to higher complication rates. Therefore, cardioversion should be considered in older patients with the same criteria and emphasis as in younger patients. Anticoagulation and antithrombotic medication is underused for cardioversion and in treating chronic AF, especially in elderly patients.