This article reports the outcome of 29 children with Reye syndrome admitted to Yale-New Haven Hospital since initiating a regimen incorporating intracranial pressure (ICP) monitoring, controlled respiration, and administration of mannitol and barbiturates. The majority of the children were in stage 3 coma and 24/29 had ICP monitored via an intraventricular catheter. Evidence of severe disease characterized by blood ammonia concentrations >500 µg/100 ml, and difficulty controlling ICP manifested by the need to administer numerous doses of mannitol were exhibited by 17 of the children, and longer hospitalizations were required for them. The earliest and most reliable indicator of severity was the degree of elevation of the blood ammonia concentration; other laboratory values (SGOT, lactic dehydrogenase, creatine phosphokinase), while related to each other, did not reliably predict severity. ICP monitoring was maintained for a median duration of eight days and ICP was most difficult to control for the first few days of the illness. Two children died during the acute illness (2/29, mortality 6.9%) and there were significant neurologic residua in three children (10.3%). The outcome using our present management technique was clearly superior to previous management at our own institution when considering the most severely affected children. Prior to ICP monitoring (1967-1974) we experienced a 60% mortality in children with blood ammonia >500 µg/100 ml; in this present series we noted two deaths and two children with significant residua in the group of 17 children with blood ammonia >500 µg/100 ml, an overall salvage rate of 11/15 (73.3%). At the present time we are not able to determine which measure or combination of measures is necessary or whether addition to or modification of the treatment protocol might be more effective. Resolution of such issues must await the results of a prospective, multi-institutional collaborative project.