Background: BNP level were influenced by various stimuli and condition in hypertensive patients. So, we evaluated that the relationship LV geometry or wall stress and BNP early hypertensive patients. Methods: We studied 42 prospectively consecutive middle-aged subjects, with newly diagnosed stage I-II essential hypertension (Group I) and 42 age and sex-matched normotensive individuals from health screening program (Group II). Office blood pressure (BP) monitoring, plasma levels of brain natriuretic peptide (BNP, Biosite), metabolic profile and an echocardiographic measurement constituted the work-up of all subjects. Exclusion criteria were overt heart failure, renal insufficiency ( 1.5), severe obesity ( BMI 30) or any other severe concomitant illness. We defined LVH as LVMI >117(men) or >104(women) g/m2 in echocardiography. Results: Baseline demographic features were similar in both groups. Incidence of LVH on ECG were not different (4.8% vs 2.4%, p=0.557) but LVH on Echo (45.2% vs. 19%, p=0.010) and Relative wall thickness (35.7% vs. 16.7%, p=0.047) were significant different. Mean BNP level was different between 2 groups (BNP; 159.1 짹 307.4 pg/ml vs 7.2 짹 9 pg/ml, p=0.013). The area under the receiver-operating curve (AUC) for initial BNP to detect LVH was 0.617(p=0.0285). A BNP value of 18.05pg/mL was associated with the most optimal sensitivity of 78.6% and a specificity of 53.3% for detecting LVH. Initial BNP level was significantly correlated with LVMI and End-systolic wall stress(ESWS)(r=0.503 & 0.465, p=0.001). In multivariate analysis, risk for LVH were age (p=0.003) and hypertension history (p=0.015). CONCLUSIONS: BNP level was related to LV geometry and stress and BNP testing for identification of LV hypertrophy patients in hypertension is useful method.