We have previously reported concordant changes in cerebral oxygenation measured by near infrared spectroscopy (NIRS) and mean arterial blood pressure(MAP) in critically ill premature infants (Tsuji et. al., Pediatr Res. 37:241A, 1995). We hypothesized that these changes are due to MAP-induced alterations in cerebral blood flow (CBF), i.e., a pressure passive circulation, and studied anesthetized neonatal piglets (n=6) during hypotensive episodes. To test this hypothesis, relative changes in cerebral oxygenation and cerebral blood volume were measured by NIRS from the hemoglobin difference signal (HbD=oxyhemoglobin-deoxyhemoglobin) and the total hemoglobin signal (HbT=oxyhemoglobin + deoxyhemoglobin). CBF velocity (TAV) in the anterior cerebral artery was measured by Doppler ultrasound, and volemic CBF was measured by the radioactive microsphere technique. Hypotension to<50% of baseline MAP was achieved by an adjustable ligature around the ascending aorta. Arterial oxygenation was maintained constant by mechanical ventilation. HbD and MAP showed concordant changes as observed in our studies of sick premature infants. HbT changes were less clearly concordant. To determine whether the decrease in cerebral intravascular oxygenation produced by hypotension related to a decrease in CBF, simultaneous measurements of CBF by the radioactive microsphere technique were made. Hypotension was shown to produce significant decreases in CBF (42.8 ± 12.5% of baseline p<0.01) and cerebral oxygenation (-14.8 ± 5.0 micromol/L p< 0.01) with a smaller, less consistent decrease in cerebral blood volume (- 2.6±3.2 micromol/L p NS). Analysis of covariance of the NIRS and Doppler data showed a significant correlation between HbD and TAV (p=0.01) and between HbD and MAP (p<0.05). HbT was less significantly correlated with TAV(p=0.03) and did not reach statistical significance versus MAP. We conclude that (1) alterations in MAP produce altered cerebral intravascular oxygenation by altering CBF and hence oxygen delivery, (2) the HbD signal is more sensitive to changes in cerebral perfusion than the HbT signal, and (3) NIRS recordings in premature infants showing concordant HbD and MAP signals imply a“pressure passive” state and may identify an infant at risk for cerebral injury.