Animal experiments have outlined deleterious effects of glutamate, an excitotoxic amino acid, in cerebral tissue following hypoxia-ischemia (H.I.). Glutamate can be demonstrated in vivo in cerebral tissue using1 H-MRS. The present study tested the hypothesis, that elevated levels of glutamate in brain tissue of neonates with cerebral H.I. predicted an adverse neurodevelopmental outcome. Methods, subjects: Studies were performed in 11 fullterm patients with global brain H.I. and in 3 preterm neonates (gestational age 32.0 ± 4.4 weeks) with localized brain H.I. Age at test was 10.4 ± 7.2 days in the fullterm patients, and term age in the preterm patients. Neurodevelopment was assessed at 3, 6, and 9 months. Neonatal brain metabolism was studied using 1H-MRS in a 1.5 T field. After MRI, MR spectra were obtained in two PRESS volumes of 16 cm3 each, examining both hemispheres. Pulse sequences included adiabatic pulses for water suppression, TR/TE: 2000/30 ms. As N-acetylaspartate (NAA) is a marker for neurons, peak-to-peak ratios of glutamate/NAA (Glx/NAA) were calculated. Results: Four fullterm patients showed an abnormal neurodevelopment (mental retardation, spastic quadriplegia), and all three preterm patients developed hemiplegia. Glx/NAA ratios in patients with global hypoxia and a normal outcome were 0.48 ± 0.11, whereas ratios in patients with H.I. and an adverse outcome were 0.59 ± 0.15. In patients with localized lesions, Glx/NAA ratios were 0.46 ± 0.19 in the normal side of the brain, and 0.57 ± 0.17 in the affected side. Although these differences did not reach significance, we conclude that Glx/NAA ratios tended to be higher in neonates with global brain H.I. and an adverse outcome and in localized H.I. brain lesions, supporting our hypothesis. As glutamate is produced in the early stages after H.I., differences between patients with a good or a poor outcome might be even more pronounced at an earlier age following cerebral hypoxia-ischemia.