We compared simultaneous measurement of aortic, direct (Dir) and of indirect (Ind) systolic (S), mean (M) and diastolic (D) arterial pressure (AP) determined by an automatic oscillometric instrument (DINAMAP) in neonates with birthweight of 1000-3680 g. DINAMAP measurements were performed with cuffs of increasing width and length, recommended by the manufacturer for increasing arm circumference, and with a Standard sized cuff (2.5 X 15 cm), previously considered as suitable for neonates of any body size. In addition, we compared simultaneous measurements of Dir SAP and of Ind SAP determined by a Doppler technique and the Standard cuff. In DINAMAP SAP measurements with the Standard cuff, a statistically significant correlation between arm circumference and delta Ind-Dir SAP values (i.e. the difference between simultaneous Ind and Dir SAP measurements) was found. In DINAMAP MAP measurements with the cuff recommended for arm circumference, a statistically significant difference of the mean delta Ind-Dir MAP values was observed in infants whose arm was or was not completely encircled by the bladder of the cuff. In SAP as well as in MAP DINAMAP determinations, the overall error of measurement with the Standard cuff was smaller than with the recommended cuff. The Doppler method was found considerably more accurate than the DINAMAP method for the determination of SAP. In spite of these limitations, the DINAMAP method with the Standard cuff was considered to be reasonably accurate for the clinical determination of SAP and MAP, provided that several consecutive measurements are performed and averaged in order to minimize the error of measurement. When considering DAP measurements the error was so unacceptably high that the DINAMAP method cannot be recommended for clinical use. The need for a careful consideration of the cuff characteristics when evaluating new methods for the indirect measurement of AP in the neonate is emphasized.