The authors studied 37 term neonates (38-42 gestational weeks) at 1-11 days after central nervous system insult to determine whether proton magnetic resonance spectroscopy (1H-MRS) of the occipital gray/parietal white matter was useful in predicting outcomes. Etiologies included asphyxia, 18; sepsis/meningitis, 8; metabolic disorders, 5; stroke, 4; and trauma, 2. 1H-MRS data (1.5T; 8 cm3 vol, stimulated echo acquisition mode sequence, TE = 20 ms, TR = 3000 ms) were expressed as metabolite peak area ratios (NAA/Cr, NAA/Cho, Cho/Cr) and the presence or absence of lactate. Outcomes were assessed at 6 to 12 months post-insult using the Pediatric Cerebral Performance Scale and were dichotomized as follows: good/moderate outcome (good, mild or moderate disability) or poor outcome (severe disability, persistent vegetative state, death). Neonates with poor outcomes had significantly lower NAA/Cho and significantly higher Cho/Cr ratios in the occipital region, as compared with patients with good/moderate outcomes. No neonates with good/moderate outcomes had metabolite ratios that exceeded 2 standard deviations from the mean. In addition, the absence of lactate on 1H-MRS correlated with a good/moderate outcome. The study also showed that 1H-MRS metabolite ratio data, added to either the Sarnat or EEG scores, enhanced the correlation between these prognostic factors and outcomes. 1H-MRS provides additional objective data early after a wide variety of perinatal neurologic insults to enhance outcome prediction.