By using proton magnetic resonance spectroscopy ((1)H-MRS), cerebral lactate has been shown to be elevated in a wide variety of pediatric and adult neurological diseases. In this study we compared 36 newborns, infants, and children with elevated lactate peaks on (1)H-MRS with 61 patients without an identifiable lactate signal. (1)H-MRS was acquired from the occipital gray and parietal white matter (8 cm3 volume, STEAM sequence with echo time = 20 msec, repetition time = 3.0 seconds) and data were expressed as ratios of different metabolite peak areas (N-acetylaspartate [NA]/creatine [Cr], NA/choline [Ch], and Ch/Cr) and the presence of a characteristic lactate doublet peak at 1.3 ppm. Outcomes (Pediatric Cerebral Performance Category Scale score; PCPCS) were assigned 6 to 12 months after injury. Patients with lactate peaks were more likely to have suffered a cardiac arrest, were more often hyperglycemic, and had lower Glasgow Coma Scale scores on admission. They were also more likely to have abnormal metabolite ratios when compared with age-matched controls or with patients without detectable lactate. Of prognostic importance, patients with increased lactate were more likely to be severely disabled (39% vs 10%), survive in a persistent vegetative state (13% vs 2%), or have died (39% vs 7%). In contrast, patients with similar conditions without increased lactate were more likely to have had a good outcome (23% vs 3%) or recovered to a mild (38% vs 6%) or moderate disability (20% vs 0%). Our data suggest that (1)H-MRS is useful in the prediction of long-term outcomes in children with neurological disorders. Patients with elevated cerebral lactate are more likely to die acutely or are at greater risk for serious long-term disability.