One cause of cerebral damage in comatose patients resuscitated from a cardiac arrest is cerebral ischaemia occurring during the postresuscitation period. Near-infrared spectroscopy has been advocated as a useful monitor of brain oxygenation, but data on clinical use in comatose postarrest patients are not available. Therefore, we compared regional oxygen saturation measured with the INVOS 3100 with global oxygen saturation measured using jugular bulb oximetry in 10 comatose patients successfully resuscitated from an out-of-hospital cardiac arrest. Our data show that, in most patients, there is a significant difference between the two methods. The INVOS 3100 over-read at low jugular bulb saturations (< 60%) and under-read at high jugular bulb saturations (> or = 60%). During hypoventilation we found a significant increase in regional oxygen saturation. This increase was significantly correlated with the rise in cerebral blood flow and cardiac index, indicating that regional oxygen saturation is influenced by both cerebral and extracerebral components. We conclude that regional cerebral oxygen saturation measured with the INVOS 3100 cannot be compared with global cerebral oxygen saturation measured with jugular bulb oximetry in comatose patients resuscitated from a cardiac arrest. This may be due to characteristics of the INVOS 3100 or to the distribution of cerebral blood flow after cardiac arrest.