The association of ventricular arrhythmias with left ventricular (LV) hypertrophy was examined in 6,218 participants in the Framingham Heart Study. Electrocardiographic (ECG) LV hypertrophy was present in 171 subjects and echocardiographic hypertrophy was detected in 869. Echocardiographic LV hypertrophy was associated with increased risk for each of 6 ventricular arrhythmia grades in men (relative risk up to 8.9, p less than 0.01), and 4 of 6 grades in women (p less than 0.05). Similarly, men with ECG LV hypertrophy were at increased risk for 4 of 6 arrhythmia grades (p less than 0.05). However, owing to low prevalence ECG LV hypertrophy was not associated with arrhythmia in women. After adjustment for age, sex, systolic blood pressure, valvular heart disease, angina pectoris and acute myocardial infarction, the association of echocardiographic but not ECG LV hypertrophy with ventricular arrhythmia remained significant (p less than 0.001). Thus, echocardiographic LV hypertrophy is more prevalent and more sensitive for ventricular arrhythmias than ECG LV hypertrophy.