ObjectiveTo investigate the prevalence and clinical determinants of atrial cardiopathy in patients with embolic stroke of unknown source (ESUS) and compare with other established stroke etiologies.MethodsIn a cross-sectional study of 846 consecutive patients with ischemic stroke, we compared the prevalence of atrial cardiopathy (defined by p-wave terminal force in V1 >5,000 µV·ms or severe left atrial enlargement) between ESUS patients and patients with large artery atherosclerosis (LAA) and small vessel disease (SVD) strokes. Baseline characteristics were also compared between ESUS and cardioembolic (CE) patients.ResultsOf all, 158 (19%) patients met ESUS diagnostic criteria, while others were classified into LAA (n = 224, 26%), SVD (n = 154, 18%), and CE (n = 310, 37%). The prevalence of atrial cardiopathy was higher in ESUS patients compared to noncardioembolic stroke patients (26.6% vs 12.1% in LAA vs 16.9% in SVD; p = 0.001). ESUS patients were younger, were less hypertensive, and had higher cholesterol and low-density lipoprotein levels, but also had less left ventricular or atrial abnormalities when compared to CE patients.ConclusionThe prevalence of atrial cardiopathy was high in ESUS patients compared with patients with nonembolic strokes. Interestingly, ESUS patients were also clinically different from CE patients. While the presence of atrial cardiopathy may reflect a unique mechanism of thromboembolism in ESUS patients, it is still unclear if they may benefit from anticoagulation, or if the presence of atrial cardiopathy in this population could serve as a risk-stratifying marker for stroke recurrence. Further efforts are necessary to provide better characterization of the ESUS population in order to develop better stroke preventive strategies.