The echocardiogram (Echo) is the validity standard for left ventricular mass (LVM) and LV hypertrophy (LVH). Numerous studies have confirmed modest correlations between the electrocardiogram (ECG) and Echo for LVM and low ECG sensitivity for Echo-LVH. In spite of this, investigators continue modeling ECG parameters to optimize their relation with the Echo. The authors studied the association between eight ECG criteria and Echo-LVM estimates in a biracial population of men and women with mild hypertension. The Echo-LVM was determined by the Penn convention and expressed in grams, g/m, g/m2, and g/m2.7. The ECGs and Echos were recorded at baseline, 3 months, and annually for 4 years. The ECGs were computer processed to define the following criteria: (1) Casale/Devereux, (2) Cornell product, (3) Cornell voltage, (4) 12-lead voltage product, (5) sum of the 12-lead, (6) Rautaharju, (7) Sokolow-Lyon, and (8) Romhilt-Estes point score. The major findings were: (1) correlations between the ECG and Echo were modest for level and minimal for change, (2) Echo indexing did not alter correlations with ECG criteria, (3) white men and women show higher correlations for level and change than blacks, (4) repeatability of the Echo-LVM index was 0.7, making it a "moving" validity standard for the ECG, (5) further ECG modeling to predict Echo-LVH, especially in whites, is not a productive approach, and (6) ECG measurements should be combined with other non-ECG characteristics when detecting LVH, and future ECG-LVM studies should investigate the prognostic value of ECG characteristics and use disease outcome as the validity standard.