Cross-brain oxygen extraction may be altered by coma, hyperventilation, hypothermia, or barbiturates, and has been demonstrated in adults and more recently in children to be related to functional neurologic recovery after a variety of brain injuries. However, measurement of cross-brain oxygen extraction in children is currently not a part of routine clinical care, partly because there have been no published attempts relating the technique of jugular venous bulb (JVB) catheterization and its complication in children. We catheterized the JVB to measure cerebral venous oxygen content and calculate cross-brain oxygen extraction in 26 deeply comatose neonates and children ranging in age from a few hours to 14 yr. Bedside catheterization using the Seldinger technique was successful in 25 children, with standard venous cutdown necessary in the remaining child. All JVB catheterizations were performed with parental consent and during continuous monitoring of the intracranial (ICP) or fontanelle, as well as arterial, pressure. ICP was not significantly altered by the cannulation procedure in any of the children studied, although the cannulation occurred early in the child's course when ICP was well controlled. Inadvertent carotid artery puncture with bleeding controlled by local pressure occurred in four children, and catheter malposition was confirmed on lateral skull xray in two others. Jugular venous bulb catheters remained in place for 2 to 7 days (average 3) and malfunction or obstruction of the catheter did not occur. Organisms were grown from three of 26 catheter tips submitted for culture, with peripheral blood cultures also positive for the same organisms in two of these.(ABSTRACT TRUNCATED AT 250 WORDS)