Although left ventricular hypertrophy (LVH) detected by electrocardiography (ECG-LVH) and echocardiography (echo-LVH) independently predict cardiovascular disease events, it is unclear if ECG-LVH and echo-LVH independently predict atrial fibrillation (AF).This analysis included 4,904 participants (40% male; 85% white) from the Cardiovascular Health Study who were free of baseline AF and major intraventricular conduction delays. ECG-LVH was defined by Minnesota Code Classification from baseline ECG data. Echo-LVH was defined by sex-specific left ventricular mass values >95th sex-specific percentiles. Incident AF events were identified during the annual study ECGs and from hospitalization discharge data. Cox regression was used to compute hazard ratios (HR) and 95% confidence intervals (CI) for the association of ECG-LVH and echo-LVH with incident AF, separately.ECG-LVH was detected in 224 (4.6%) participants and echo-LVH was present in 231 (4.7%) participants. Over a median follow-up of 11.9 years, a total of 1,430 AF events were detected. In a multivariable Cox model adjusted for age, sex, race, education, income, smoking, systolic blood pressure, diabetes, body mass index, total cholesterol, high-density lipoprotein cholesterol, aspirin, antihypertensive medications, and cardiovascular disease, ECG-LVH (HR = 1.50; 95% CI = 1.18, 1.90) and echo-LVH (HR = 1.39; 95% CI = 1.09, 1.78) were independently associated with AF. When ECG-LVH (HR = 1.47, 95% CI = 1.16, 1.87) and echo-LVH (HR = 1.36, 1.07, 1.75) were included in the same model, both were predictive of incident AF.The association of ECG-LVH with AF is not dependent on left ventricular mass detected by echocardiography, suggesting that abnormalities in cardiac electrophysiology provide a distinct profile in the prediction of AF.