Blood pressure (BP) measurement in clinical assessment by means of a mercury sphygmomanometer (MS) has numerous drawbacks. It has been proposed that non-invasive, 24-hr ambulatory blood pressure monitoring (NIABPM) should provide more appropriate BP values for both the diagnosis of hypertension and for its subsequent monitoring during treatment. The aim of the present study was to investigate, in 100 ambulatory and 250 hospitalized elderly subjects, the prevalence of white coat hypertension (WCH) and masked hypertension (MH) in a cohort of older subjects, by using both clinical readings (MS) and NIABPM. The results of our analyses indicate that a higher prevalence of old-old subjects regarded as normotensive at anamnesis or with normal BP values at MS are true hypertensive (50% of ambulatory patients and 17.1% of hospitalized patients) or masked hypertensive (10.3% of ambulatory patients and 28.6% of hospitalized patients), and consequently at higher risk of cardiovascular morbidity and mortality.