Background Angiotensin II receptor blockers (ARB) are now commonly used to treat hypertension because of their beneficial effects on cardiovascular remodeling. However, ARB treatment can not inhibit the left ventricular (LV) remodeling sufficiently, which may be related with aldosterone secretion. To inhibit the action of aldosterone during ARB treatment, the additional effects of an aldosterone blocker and spironolactone (SPRL) on LV hypertrophy in patients with essential hypertension was studied. Methods and Results The patients with essential hypertension were randomly divided into 2 groups; 1 group was treated with an ARB, candesartan (8 mg/day), for 1 year (ARB group) and other group was treated with the ARB for the first 6 months and with the ARB plus SPRL (25 mg/day) for the next 6 months (combination group). Seventy patients who underwent echocardiography every 6 months were analyzed and were also classified into 4 subgroups of LV geometric pattern according to the LV mass index (LVMI) and the relative wall thickness (RWT). The ARB treatment and the addition of SPRL significantly reduced the blood pressure, however, both treatments did not affect the LV geometry in both groups. The ARB treatment in the subgroups of concentric LV remodeling (RWT ≥0.45 and LVMI <125) and concentric LV hypertrophy (RWT ≥0.45 and LVMI ≥125) significantly reduced RWT. However, ARB treatment in all subgroups did not affect LVMI. The addition of SPRL only in the concentric LV hypertrophy subgroup significantly reduced the LVMI, despite similar changes in blood pressure. Conclusions These results indicated that the addition of SPRL treatment during the ARB treatment and conventional treatments is clinically useful to reduce the LVMI in hypertensive patients with concentric LV hypertrophy; however, does not improve the eccentric LV hypertrophy. (Circ J 2006; 70: 995 - 1000)