In essential hypertension, the regression of left ventricular hypertrophy is an important goal of treatment. In addition to treatment-associated changes in blood pressure (BP), the roles of other determinants of left ventricular hypertrophy regression, including dietary sodium intake, deserve investigation. In the present study, the change in echographic left ventricular mass index (LVMI) was assessed in 182 patients with never-treated essential hypertension at baseline and after 3 years of treatment by angiotensin converting enzyme inhibitors or angiotensin II receptor antagonists given at recommended doses and associated with other antihypertensive agents. Treatment was associated with satisfactory control of BP (<140/90 mm Hg) in 64% of patients, and left ventricular hypertrophy prevalence decreased from 56% to 39%. Twenty-four-hour urinary sodium was positively related to LVMI at baseline and at the end of the study, independently of age, sex, and systolic BP. Supine plasma aldosterone concentration was correlated with LVMI only at baseline but not in multivariate analysis. In response to treatment, the percentage of change in LVMI was positively correlated with the absolute changes in systolic BP, urinary sodium, and plasma aldosterone concentration, independently of baseline LVMI. The population was divided into 3 tertiles according to final values of gender-specific urinary sodium. When, within each urinary sodium tertile, patients were divided into those with plasma aldosterone concentration below and above the median (11.6 ng/dL), LVMI progressively increased across sodium tertiles only in patients with high plasma aldosterone concentration. Systolic BP was similar across all of the groups. In conclusion, aldosterone requires the presence of high dietary salt for the expression of its unfavorable effect on the heart.