Long-term effects of antihypertensive treatment on various electrocardiogram (ECG) voltages and the association between ECG findings at randomization and subsequent mortality were evaluated in the double-blind, placebo-controlled trial of elderly hypertensive patients, conducted by the European Working Party on High Blood Pressure in the Elderly (EWPHE). Patients were treated with a combination of hydrochlorothiazide and triamterene or matching placebo; methyldopa or placebo was added if blood pressures remained high. RaVL and SV1 + RV5 at randomization were related to systolic blood pressure (SBP) and RaVL was also related to diastolic blood pressure (DBP), after adjustment for age, gender, and body mass index (BMI). When adjusted for age and body mass index, the decreases in RaVL and SV1 + RV5 were not correlated with the changes in SBP after 1 year of active treatment but the decreases in SV1 + RV5 were positively related to the fall in DBP. After 4 years of active treatment, the fall in SBP was associated with decreases in RaVL and in SV1 + RV5, after adjustment for age and changes in BMI.