ObjectiveTo compare the predictive powers of self-measurement of blood pressure at home (home blood pressure measurement) and casual (screening) blood pressure measurement for mortality.DesignA prospective cohort study.Subjects and methodsWe obtained home and screening blood pressure measurements for 1789 subjects aged ≥ 40 years who were followed up for a mean of 6.6 years. The prognostic significance of blood pressure for mortality was determined by the Cox proportional hazards regression model adjusted for age, sex, smoking status, past history of cardiovascular disease, and the use of antihypertensive medication.ResultsWhen the home blood pressure values and the screening blood pressure values were simultaneously incorporated into the Cox model as continuous variables, only the average of multiple (taken more than three times) home systolic blood pressure values was significantly and strongly related to the cardiovascular mortality risk. The average of the two initial home blood pressure values was also better related to the mortality risk than were the screening blood pressure values.ConclusionsHome blood pressure measurement had a stronger predictive power for mortality than did screening blood pressure measurement for a general population. This appears to be the first study in which the prognostic significances of home and screening blood pressure measurements have been compared.