BACKGROUND: Automatic noninvasive blood pressure measurement devices should be clinically validated prior to marketing using standard criteria such as those published by the Association for the Advancement of Medical Instrumentation and the British Hypertension Society. OBJECTIVE: To assess a new bedside oscillometric device (Accutorr Plus, Datascope Inc., Montvale, New Jersey, USA). METHODS: We compared measurements obtained using the device with mercury column measurements for 90 subjects (448 measurements). In addition to calculation of the limits of agreement for the device versus observers, we determined the impact of age, the blood pressure level, the body mass, and the arm size on the levels of agreement. RESULTS: There was a good agreement between observers, with a mean difference of 0.38 +/- 4.47 mmHg for the systolic and 0.0 +/- 3.7 mmHg for the diastolic blood pressure. The observer-device agreement was -0.04 +/- 7.93 mmHg for the systolic and 0.35 +/- 5.75 mmHg for the diastolic blood pressure. Regression analyses showed that, as the systolic blood pressure and age increased, the level of disagreement for the systolic blood pressure also increased ( P < 0.001 for both). CONCLUSIONS: The Accutorr Plus satisfied the criteria of the Association for the Advancement of Medical Instrumentation for a general adult population with large ranges in age, blood pressure, body mass and arm circumference. Not unlike many oscillometric devices, for systolic blood pressures > 190 mmHg, the device error increased.