This study evaluates the outcome of 56 severely brain injured children (mean age 6.2±2.1 years) and relates the Initial Glasgow Coma Scale (IGCS), initial intracranial pressure (ICP int), maximal intracranial pressure (ICP max) and minimal cerebral perfusion pressure (CPP min) to quality of survival. Forty-one children sustained head trauma, five severe central nervous system infections and 10 were of miscellaneous etiology. Therapy consisted of mechanical hyperventilation, moderate fluid restriction, dexamethasone and diagnosis specific measures when indicated. Outcome was categorized according to the Glasgow outcome scale at discharge from the hospital. An IGCS of 3 was associated with 100% mortality, 7 and above resulted in 72% good recovery, 28% poor outcome and no mortality. ICP int of less than 20 torr was noted in (67%) of the patients, and did not correlate with ICP max or outcome. Conversely, ICP int in excess of 40 torr correlated well with ICP max and outcome. ICP max of less than 20 torr resulted in 57% good recovery, 36% poor outcome and 7% mortality. ICP max greater than 40 torr resulted in 7% poor outcome and 93% mortality (p<0.001). In head trauma, 32 patients (78%) were alive with mean ICP max 16.9±3.1 and CPP min 65.5±8.5 torr compared to 9 patients (22%) who died with mean ICP max 53.7±10.8 and CPP min 6±3.9 torr, (p<0.01). In children with infectious etiology 60% survived with mean ICP max 16±3 and CPP min 96±16 torr. Forty percent died with ICP max 39±21 and CPP min 65±35 torr. In patients with brain injury of miscellaneous etiology, 60% lived with mean ICP max 17.3±6.2 and CPP min 75±13.4 torr, 40% died with mean ICP max 45.6±7.8 and CPP min 23.3±9.4 torr. Our results indicate favorable outcome in head trauma patients and stress the overall importance of aggressive ICP control and optimal maintenance of cerebral perfusion pressure. The same does not imply in CNS infection, since a mean CPP min of 65±35 torr in non survivors did not affect outcome.