Cerebral blood flow pressure-passivity results when pressure autoregulation is impaired, or overwhelmed, and is thought to underlie cerebrovascular injury in the premature infant. Earlier bedside observations suggested that transient periods of cerebral pressure-passivity occurred in premature infants. However, these transient events cannot be detected reliably by intermittent static measurements of pressure autoregulation. We therefore used continuous bedside recordings of mean arterial pressure (MAP; from an indwelling arterial catheter) and cerebral perfusion [using the near-infrared spectroscopy (NIRS) Hb difference (HbD) signal) to detect cerebral pressure-passivity in the first 5 d after birth in infants with birth weight <1500 g. Because the Hb difference (HbD) signal [HbD = oxyhemoglobin (HbO2) - Hb] correlates with cerebral blood flow (CBF), we used coherence between MAP and HbD to define pressure-passivity. We measured the prevalence of pressure-passivity using a pressure-passive index (PPI), defined as the percentage of 10-min epochs with significant low-frequency coherence between the MAP and HbD signals. Pressure-passivity occurred in 87 of 90 premature infants, with a mean PPI of 20.3%. Cerebral pressure-passivity was significantly associated with low gestational age and birth weight, systemic hypotension, and maternal hemodynamic factors, but not with markers of maternal infection. Future studies using consistent serial brain imaging are needed to define the relationship between PPI and cerebrovascular injury in the sick premature infant.