Selective head cooling has been proposed as a neuroprotective intervention after hypoxia-ischemia in which the brain is cooled without subjecting the rest of the body to significant hypothermia, thus minimizing adverse systemic effects. There are little data showing it is possible to cool the brain more than the body. We have therefore applied selective head cooling to our hypoxia-ischemia piglet model to establish whether it is possible. Nine piglets were anesthetized, and brain temperature was measured at the surface and in the superficial (0.2 cm) and deep (1.7-2.0 cm) gray matter. Rectal (6-cm depth), skin, and scalp temperatures (T) were recorded continuously. Lowering T-rectal from normothermia (39 degrees C) to hypothermia (33.5-33.8 degrees C) using a head cap perfused with cold (6-24 degrees C) water was undertaken for up to 6 h. To assess the impact of the 45-min hypoxia-ischemia insult on the effectiveness of selective head cooling, four piglets were cooled both before and after the insult, and four, only afterward. During selective head cooling, it was possible to achieve a lower T-deep brain than T-rectal in all animals both before and after hypoxia. However, this was only possible when overhead body heating was used. The T-rectal to T-deep brain gradient was significantly smaller after the insult (median, 5.3 degrees C; range, 4.2-8.5 degrees C versus 3.0 degrees C; 1.7-7.4 degrees C; p = 0.008). During rewarming to normothermia, the gradient was maintained at 4.5 degrees C. We report for the first time a study, which by direct measurement of deep intracerebral temperatures, validates the cooling cap as an effective method of selective brain cooling in a newborn animal hypoxia-ischemia model.