BACKGROUND: Blood pressure measurements with a mercury sphygmomanometer serve as the screening test to establish the clinical diagnosis of primary hypertension. But one of the problems of hypertension screening is the variability of blood pressure measurements. METHODS: In order to identify the measurement error of blood pressure through the biennial health screening program of Korea Insurance Medical Corporation, we compared the blood pressure of screening test with true value which defined as the mean of blood pressures taken from the direct contact and twice check-up among normotensive Seoul Cohort participants. RESULTS: Three hundred forty-nine participants had both data of screening measurement and survey of direct contact. The means of difference in systolic and diastolic blood pressure between both data were 8.47 mmHg and 6.54 mmHg, respectively. These results showed statistical significance with paired t-test (p=0.001). CONCLUSION: Our findings indicated that screening measurement of blood pressure had false negative. And if a epidemiologic study about blood pressure used the data of screening test, it should adjust value of systolic and diastolic blood pressure with adding 8 and 7 mmHg, respectively.