To determine the independent prognostic significance of left ventricular (LV) mass and geometry (concentric vs eccentric pattern) in hypertensive subjects with LV hypertrophy at echocardiography, 274 subjects were followed for up to 8.7 years (mean 3.2). All patients had systemic hypertension and LV mass > or = 125 g/body surface area (BSA) and underwent ambulatory blood pressure (BP) monitoring and echocardiography before treatment. Eccentric and concentric hypertrophy were defined by the ratio between LV posterior wall thickness and LV radius at telediastole <0.45 and > or = 0.45, respectively. Age, sex ratio, body mass index, office BP and serum glucose, cholesterol, and triglycerides did not differ between the groups with eccentric (n=145) and concentric (n=129) hypertrophy. Average 24-hour daytime, and nighttime systolic ambulatory BPs were higher in concentric than in eccentric hypertrophy (all p <0.01). LV mass was slightly greater in concentric than in eccentric hypertrophy (157 vs 149 g/BSA, p <0.05). Endocardial and midwall shortening fraction were lower in concentric than in eccentric hypertrophy (96.5% vs 106.0% of predicted and 71.4% vs 89.7% of predicted, respectively; both p <0.01). The rate of major cardiovascular morbid events was 2.20 and 3.34 per 100 patient-years in eccentric and concentric hypertrophy, respectively (log rank test, p=NS). Age >60 and LV mass above median (145 g/BSA) were significant adverse prognostic predictors, while LV geometry (eccentric vs concentric hypertrophy) and ambulatory BP were not. The event rates per 100 patient-years were 1.38 and 3.98, respectively, in the patients with LV mass below and above median (age-adjusted relative risk 2.70; 95% confidence interval [CI] 1.03 to 6.63; p=0.015). In hypertensive subjects with established LV hypertrophy, LV mass, but not its geometric pattern, provides important prognostic information independent of conventional risk markers including office and ambulatory BP.