Cerebrovascular pressure autoregulation (CPA) regulates cerebral blood flow (CBF) in relation to changes in mean arterial blood pressure (MAP). Identification of a pressure-passive cerebral perfusion and the potentially modifiable physiologic factors underlying it has been difficult to achieve in sick infants. We previously validated the near-infrared spectroscopy-derived hemoglobin difference (HbD) signal (cerebral oxyhemoglobin - deoxyhemoglobin) as a reliable measure of changes in CBF in animal models. We now sought to determine whether continuous measurements of DeltaHbD would correlate to middle cerebral artery flow velocity (CBFV), allow identification and quantification of pressure-passive state, and help to delineate potentially modifiable factors. We enrolled 43 infants (2 d to 7 mo old) who were undergoing open cardiac surgery and cardiopulmonary bypass. At 6 and 20 h after surgery, we measured changes in HbD, CBFV (by transcranial Doppler), and MAP at different end-tidal CO(2) levels. We assigned a pressure-passive index (PPI) to each study on the basis of the relative duration of significant coherence between DeltaMAP and DeltaHbD. We found a significant relationship between DeltaHbD and DeltaCBFV at both time points. At 6 h after surgery, we showed high concordance (coherence > 0.5; PPI > or = 41%) between DeltaMAP and DeltaHbD, consistent with disturbed CPA in 13% of infants. End-tidal CO(2) values > or = 40 mm Hg and higher MAP variability both were associated with increased odds (p < 0.001) of autoregulatory failure. This approach provides a means to identify and quantify disturbances of CPA. High CO(2) levels and fluctuating MAP are two important preventable factors associated with disturbed CPA.