Left ventricular mass sometimes decreases during treatment of hypertension but the effects are inconsistent and the response to long-term treatment is unknown. Therefore the long-term effects of antihypertensive therapy on echocardiographically-proven left ventricular hypertrophy (LVH) were prospectively investigated in 117 previously untreated hypertensive patients (mean age 46 +/- 9 years; 15 women and 102 men). Twenty-two patients received 100 mg of gallopamil daily, 24 patients received 200 mg of metoprolol, 35 patients received both 50 mg of atenolol and 20 mg of nifedipine (follow-up five years), 14 patients received 200 mg of acebutol plus 20 mg of nifedipine, and 21 patients received 50 mg of atenolol and 10 mg of enalapril daily (follow-up of four years). For the entire population, there was a significant (P < 0.001) decrease in left ventricular mass index (LVMI; measurements were performed blind by two observers) of 24.5% after one year, with a further continuous and significant (P < 0.001) reduction of 44.1% after five years of treatment. There was a significant (r = 0.61, P < 0.001) correlation between the extent of LVMI before therapy and the percentage of regression of LVMI after five years of treatment. In 82% of the patients, almost complete regression of LVH was achieved. Nevertheless, there was no significant change in ventricular end-diastolic dimension but fractional shortening increased by 16% (P < 0.001). It can be concluded that achieving maximum regression of LVH by antihypertensive therapy in previously untreated hypertensives takes a long time and depends not only on the drug regimen chosen but especially on the duration of treatment and the extent of pretreatment left ventricular mass.