Recently, several investigators have reported that plasma atrial natriuretic peptide (ANP) levels are elevated in patients with hypertrophic cardiomyopathy (HC).1‐3 An increased ANP level has been associated with poor prognosis in patients with heart failure. 4‐6 Moreover, the relation between plasma ANP levels and atrial fibrillation (AF) has been observed in patients with heart failure, and these levels are elevated in patients with AF.7‐9 However, in patients with HC, the relation between plasma ANP levels and AF is unclear. The purpose of this report is (1) to assess the relation between plasma ANP levels and AF, and (2) to elucidate the prognostic significance of plasma ANP levels in patients with HC. ••• Fifty-two consecutively studied patients with HC were evaluated. The diagnosis of HC was based on the echocardiographic demonstration of a hypertrophied nondilated left ventricle without hypertension and other cardiac diseases. Three patients with left ventricular systolic dysfunction, 2 patients with a short follow-up period (,6 months), and 1 patient with severe mitral regurgitation due to chordal rupture were excluded. Finally 46 patients with HC were enrolled in this study. Echocardiographic studies (Toshiba SSH 160A, Toshiba Co., Tokyo, Japan) were performed using 3.5- and 2.75-MHz transducers. The left ventricular wall thickness and dimensions were measured just below the plane of the mitral valve leaflet. The intraventricular pressure gradient was also estimated with continuous-wave Doppler. Blood samples were obtained at a clinically stable period. Peripheral venous blood samples were obtained carefully and slowly from the antecubital vein after a supine position of at least 15 minutes. Patients fasted overnight and all drugs were continued. Plasma norepinephrine and ANP levels were measured. Plasma ANP levels were measured as previously reported with a specific immunoradioactive assay using a commercially available kit (Shionoria, Osaka, Japan). 5 Follow-up was performed from the time of blood sampling using available medical records and telephone interviews with patients and/or referring physicians. The specified end points were sudden death including resuscitated ventricular fibrillation, heart failure, and ischemic stroke and/or peripheral embolism. During a mean follow-up period of 2.1 6 0.9 years, 8 cardiovascular events occurred including sudden death in 3, heart failure in 1, ischemic stroke in 3, and peripheral embolism in 1. The mean age in the 3 patients who died suddenly was 68 6 6 years (2 were women). All 3 patients died at rest without any relation to exercise. Although 1 patient had a family history of HC and sudden death, the other 2 patients did not have any accepted risk factors for sudden death. All 3 patients with ischemic stroke had complications with AF and a large infarction documented by computer tomographic scan that was thought to be due to embolism, although clinical detail was unclear in 1 patient. Data were presented as mean 6 SD. Statistical