To assess whether electrocardiographic variables are useful to detect the regression of left ventricular (LV) mass after long-term antihypertensive treatment, we related electrocardiographic voltages to echocardiographic variables before and after treatment with an ACE inhibitor, temocapril (TEM), or direct vasodilator, cadralazine (CAD). Twenty-one patients with essential hypertension were treated with either TEM (n = 11) or CAD (n = 10) for one year. LV mass index (LVMI) by echocardiography and Sokolow-Lyon voltage (SV1 + RV5), Cornell voltage (RaVL + SV3) and RV5 + RV6 by standard 12-lead electrocardiographic voltages were determined before and after treatment. Both drugs decreased blood pressure to the same extent. Both Sokolow-Lyon voltage and RV5 + RV6 tended to decrease in the ACE group (40.0 +/- 9.4 to 37.2 +/- 9.4 mm and 44.7 +/- 13.5 to 41.7 +/- 11.7 mm, respectively, N.S.), but not in the CAD group (38.4 +/- 6.8 to 39.7 +/- 7.7 mm and 42.9 +/- 10.4 to 46.8 +/- 11.2 mm, respectively, N.S.). LVMI decreased in the ACE group (-24 +/- 22 g/m2), whereas it increased in the CAD group (37 +/- 27 g/m2, p < 0.01). Change in LVMI was correlated with the changes in RV5 + RV6 and Sokolow-Lyon voltage (r = 0.73, p < 0.01 and r = 0.70, p < 0.01, respectively), but not with that in Cornell voltage. These results indicated that the changes in voltage criteria of RV5 + RV6 and Sokolow-Lyon are useful to assess the change in LVM after antihypertensive treatment in patients with essential hypertension although voltage variables in electrocardiogram were not sensitive to detect changes in LVMI.