We have previously shown that in neonates regional cerebral blood flow(rCBF) less that 10 ml/100g/min. is compatible with normal neurologic outcome on short term follow-up. To assess the long-term prognostic value of(rCBF) measured by positron emission tomography (PET) among infants with hypoxic-ischemic encephalopathy(HIE), we retrospectively reviewed in a blinded fashion all infants with HIE born between January, 1983, and November, 1989(n=28), who underwent quantitative PET measurements of rCBF in the neonatal period. The diagnosis of HIE required evidence of featal distress and neonatal neurologic syndrome. Information was obtained by chart review, detailed neurologic examination, and Stanford-Binet intelligence testing with age adjusted norms. All living subjects(n=17) were evaluated: 8 subjects had normal neurologic examinations, while 9 were abnormal. The mean neonatal rCBF in those with normal neurologic outcomes was lower than in those with abnormal neurologic examinations (18.3±8.6 vs. 32.7±14.1 ml/100gm/min, t=2.58, p=0.22). Among the 12 subjects who underwent Stanford-Binet intelligence quotient testing (ages 5-12 years), we found a strong inverse correlation between neonatal rCBF and intelligence quotient (r=-0.72, p=0.008). We conclude that 52% of the variability (r2=0.52) in school age intelligence can be explained by differences in rCBF measured in the neonatal period and that low neonatal rCBF is predictive of good neurologic outcome in survivors of HIE. We speculate that higher rCBF associated with poor outcome may reflect loss of cerebral autoregulation or other vascular regulatory mechanisms due to hypoxic-ischemic brain damage.