In this study, a recently marketed proprietary finger blood pressure monitor, the Marshall, Astro F-88, was compared with the standard auscultative brachial mercury sphygmomanometer on 125 subjects. Measurements were undertaken according to the standards set by the American Heart Association. Sensitivity of the finger blood pressure measurement was 76% for systolic and 75% for diastolic blood pressure in diagnosis of high blood pressure (systolic greater than 140 mm Hg and diastolic greater than 90 mm Hg). Specificity was 86% for systolic and 82% for diastolic blood pressure. Positive predictive values were 58% for systolic and 38% for diastolic blood pressure in the study population in which prevalence of hypertension was 12%. The correlation coefficient (Pearson) for systolic values between devices was 0.76 (P less than .0001) and 0.57 (P less than .0001) for diastolic pressure. Values obtained by the finger monitor were found to be higher than those obtained by the mercury sphygmomanometer. Mean differences and standard deviations (paired t test) for systolic and diastolic pressures between the two devices were 2.3 +/- 14.9 mm Hg (P less than .08) and 2.9 +/- 14.5 mm Hg (P less than .02), respectively. These values are not in accordance with the proposed national standards because only 48% of the systolic and 37% of the diastolic blood pressure measurements were within 5 mm Hg of the mercury sphygmomanometer measurements. Therefore, although these differences may well be due to different techniques of monitoring employed by the devices, this device is not recommended for evaluation of blood pressure.