1. The sensitivity and specificity of four sets of electrocardiographic criteria for detection of left ventricular hypertrophy were evaluated in an echocardiographic study of 100 hypertensive patients.
2. All criteria gave reasonable specificity (87–94%) but poor sensitivity (39–52%).
3. When non-obese and obese patients were studied separately, criteria based on chest lead voltages were more sensitive than limb lead criteria for detection of left ventricular hypertrophy in non-obese subjects; however, the reverse was true in obese hypertensive patients, where criteria based on limb lead voltages were more sensitive than chest lead voltage criteria.
4. These data suggest that stratification of subjects by body build might improve the diagnostic performance of the electrocardiogram for detection of left ventricular hypertrophy.