OBJECTIVE Our objective was to test the accuracy of noninvasive recordings of finger arterial pressure (FAP) using the Ohmeda Finapres (Ohmeda Monitoring Systems, Englewood, CO). METHODS Twenty patients, aged 20 to 78 years, requiring admission to the intensive care unit and placement of intraarterial catheters participated in the study. Systolic and diastolic pressures were derived from 1-hr recordings of beat-to-beat FAP and from ipsilaterally recorded intraarterial pressure (IAP) signals. In all 20 cases, we analyzed beat-to-beat discrepancies between the actual magnitude of FAP and IAP, as well as the distribution of the consecutive differences within each of the two signals. In 10 cases, spectral analysis of the frequency content of both signals was performed. RESULTS The average systolic FAP (128.1 +/- 22.4 mm Hg) did not differ from IAP (127.1 +/- 19.7 mm Hg), whereas diastolic FAP (78.1 +/- 11.9 mm Hg) was greater (71.5 +/- 10.3 mm Hg) (p < 0.001). No differences in the linear trends of FAP and IAP were observed. Overall, systolic FAP and IAP were discrepant by 0.84 +/- 13.3 mm Hg (-21.82 to 25.8 mm Hg); diastolic FAP and IAP were discrepant by 6.67 +/- 5.23 mm Hg (2.68 to 13.05 mm Hg). Despite discrepancies in the magnitude of the two signals, the contour of IAP approximated that of FAP. Spectral analysis demonstrated good reproducibility and coherence between diastolic IAP and FAP fluctuations in both low-frequency (0.01 to 0.15 Hz) and high-frequency (0.15 to 0.33 Hz) bands. The low-frequency fluctuations in FAP systolic pressure were significantly amplified (p < 0.001) (gain 1.75), whereas the high-frequency fluctuations were not. CONCLUSIONS Over the course of 1 hr, FAP followed the contour and frequency content of IAP.