A noninvasive blood pressure monitor (Finapres) that uses the methodology of Penaz to continuously display the arterial waveform from the finger has been introduced recently. The Finapres monitor overestimated systolic pressure by 5.8 +/- 11.9 mm Hg, while the Dinamap monitor underestimated systolic pressure by -6.9 +/- 9.2 mm Hg (P = 0.003). Dinamap mean and diastolic pressure biases were less than 2 mm Hg, while the Finapres biases for these variables were significantly greater (7.7 +/- 10.0 and 8.2 +/- 9.8 mm Hg, respectively). There was no difference in systolic or mean pressure precision between the two devices (approximately 10 mm Hg), but the diastolic precision of the Dinamap unit was superior to that of the Finapres. While in most patients the Finapres monitor provided continuous blood pressure data equivalent to the data from the radial artery, marked bias (greater than 15 mm Hg) was exhibited in 2 patients for all three pressure variables. Despite this bias, blood pressure changes were tracked closely in these 2 patients. We conclude that, in its current form, the Finapres monitor cannot be relied upon independently to accurately measure blood pressure in patients undergoing general anesthesia. Since the Dinamap monitor measures mean pressure reliably and accurately, we suggest that mean blood pressure values between the Finapres and Dinamap monitors be compared to guide one in interpreting Finapres data.