ObjectiveAt present, non-invasive continuous monitoring of finger arterial blood pressure by the volume-clamp technique is considered the best approach to obtain reliable assessments of beat-to-beat blood pressure. However, data on the reproducibility (accuracy and precision) of prolonged recordings and of the hemodynamics derived from wave-form analysis are not available.DesignTen patients with untreated essential hypertension and eight normotensive subjects were monitored by Portapres over 24 h in the hospital on two occasions with 1–4 weeks in-between. Physical and mental activities were standardized as far as possible to minimize intra- and intersubject biological variability. Stroke volume was obtained by the Modelflow method. Differences between the two recordings were computed separately for the day (0700 to 2300 h) and the night (2300 to 0700 h) and for all hours. Differences in stroke volume were calculated as percentage change from the first recording.ResultsAccuracy was good in both groups and bias was close to zero. Precision was also remarkable in the daytime, and at least as good as values reported in studies that used the standard intra-arterial recording. The SD of the differences in systolic and diastolic pressure in the hypertensives in the daytime were 6.6 and 4.7 mmHg, respectively. At night, precision was less good, possibly because of the 30 min finger-cuff switching: 12.5 and 6.5 mmHg for systolic and diastolic pressure, respectively. The average stroke volume did not change more than 8% at most between the first and the second recordings.ConclusionThese results indicate that the Finapres and Portapres devices are a reliable substitute for intra-arterial recording, and are most useful instruments for the study of blood pressure regulation.