We recorded finger arterial blood pressure (FINAP) in 50 male patients during various types of surgical operations. Three different types of cuffs were used on four fingers of each patient. Measurements were made by the arterial volume-clamp method of Penaz. The FINAP measurements were compared with pressure data obtained ipsilaterally from a radial artery catheter-transducer system (intraarterial pressure [IAP]) to find optimal recording conditions and to document factors affecting FINAP readings. The thumb, with a specially designed cuff, gave the most accurate results. The mean FINAP - IAP difference for the thumb was -4.8 mm Hg for systolic pressure, 1.49 mm Hg for diastolic pressure, and 0.29 mm Hg for mean pressure. The differences were statistically significant for systolic and diastolic pressure but not for mean pressure. The regression slope for thumb systolic FINAP/IAP was 0.979, that for thumb diastolic FINAP/IAP was 0.963, and that for mean thumb FINAP/IAP was 0.996, whereas the intercepts were 7.499 for systolic pressure, 0.802 for diastolic pressure, and 0.083 for mean pressure. The correlation coefficients were 0.945 (systolic), 0.884 (diastolic), and 0.949 (mean). The correlation coefficients with the other fingers ranged from 0.502 to 0.922 for systolic pressure, 0.757 to 0.932 for diastolic pressure, and 0.767 to 0.892 for mean pressure. The slopes for the various finger-cuff combinations ranged from 0.537 to 0.996, and the intercepts ranged from 0.083 to 32.387 from mean pressure. In 3 patients (6%) the FINAP measurement was not possible because of insufficient peripheral circulation. In 9 other patients (18%) the FINAP measurements were not accurate during some periods of time.(ABSTRACT TRUNCATED AT 250 WORDS)