Anticoagulant therapy is not conventionally used in the treatment of patients with atrial flutter. This recommendation has been based on sparse clinical experience, and recent preliminary reports suggest a significant risk of thromboembolism for these patients. A retrospective study was undertaken to assess the frequency of thromboembolic events as well as potential risk factors for these events in a cohort of patients with atrial flutter referred for radiofrequency ablation treatment. Eighty-six consecutive patients with a primary diagnosis of atrial flutter were evaluated. A history of embolic events was noted in 12 of 86 patients (14%) with atrial flutter, with an annual risk of approximately 3%. There were no differences in the prevalence of coronary artery disease, cardiomyopathy, valvular disease, or atrial fibrillation between the 2 groups of patients having an embolic event and those of patients without embolic events. Left ventricular function and left atrial size were also similar between the 2 groups. The only significant risk factor was hypertension (p < 0.05). However, in a regression model with other clinical variables (i.e., age, gender, left atrial size, presence or absence of any cardiac disease, length of time in flutter, left ventricular function, type of flutter, flutter cycle length, type of secondary arrhythmias) no significant predictors were found. Patients with transient ischemic attacks or pulmonary emboli were then excluded from the analysis in order to compare the thromboembolic risk in the present study to that reported in major atrial fibrillation trials. The overall risk becomes 7% (6 of 86), which over a mean follow-up period of 4.5 years yields an annual risk of approximately 1.6%. Although this risk is only 1/3 of that for patients with atrial fibrillation, this risk is higher than previously recognized for patients with chronic atrial flutter. Anticoagulant therapy should be seriously considered for these patients.