OBJECTIVE Primarily to investigate the long-term effects of antihypertensive therapy on left ventricular morphology in non-malignant essential hypertension and in particular to compare an angiotensin converting enzyme inhibitor and a diuretic in this respect. METHODS Previously untreated males aged 20-65 years with diastolic blood pressure > or = 95 mmHg during a 4- to 6-week placebo period were randomly assigned to double-blind treatment with enalapril or hydrochlorothiazide. METHODS Indirect and intra-arterial blood pressure, echocardiography, apex cardiography, carotid pulse tracing and phonocardiography. RESULTS Left ventricular mass (LVM) was significantly correlated with intra-arterial blood pressure at baseline. During long-term treatment (14-18 months) blood pressure decreased significantly in both treatment groups (indirectly and intra-arterially), at rest and during dynamic exercise. No significant differences in blood pressure response were seen between the two therapeutic alternatives. LVM decreased progressively and significantly on enalapril after 18 months of monotherapy and decreased non-significantly on hydrochlorothiazide after 14 months of monotherapy. The difference in effect between treatments was significant in a stepwise regression analysis taking change in blood pressure into account. The relationship between the reductions in LVM and blood pressure were significant for enalapril but not for hydrochlorothiazide. Neither therapy affected left ventricular diastolic or systolic diameters significantly, but enalapril reduced posterior wall thickness and interventricular septal thickness significantly, and improved left ventricular distensibility significantly. Neither therapy had any negative effects on systolic function, although hydrochlorothiazide decreased left ventricular ejection time index significantly. CONCLUSIONS Enalapril was significantly more effective than hydrochlorothiazide in reversing left ventricular hypertrophy without negatively affecting left ventricular function.