Based on earlier structural and functional neuroimaging studies, we specifically wanted to assess N-acetylaspartate (NAA), choline-containing compounds (CHO), and creatine+phosphocreatine (CRE) levels in brain hippocampus previously demonstrated to be involved in the pathophysiology of bipolar disorder which have not been evaluated in first-episode patients. Twelve patients meeting DSM-IV criteria for bipolar disorder who consecutively applied to our department and 12 healthy controls were studied. The patients and controls underwent proton magnetic resonance spectroscopy ((1)H MRS), and measures of NAA, CHO, and CRE in hippocampal regions were obtained. ANOVA revealed in the hippocampus a significant effect of diagnosis for NAA/CRE and for NAA/CHO but not for CHO/CRE. Post hoc analysis showed that patients had a significant bilateral reduction of NAA/CRE and of NAA/CHO. No significant correlation was found between hippocampus volume and ratio measures. Correlation analyses exhibited significant correlation between NAA values and the YMRS for both side of the hippocampus, but not any other clinical variables (age, age at onset, and duration of illness). In summary, hippocampal neuronal abnormalities seem to be present at the onset of bipolar I disorder. These data suggest that neuronal abnormalities in hippocampus may be associated with the severity of bipolar I disorder. As these data were obtained in patients in their first-episode (all the patients were manic), they cannot be explained by chronicity of illness or pharmacological treatment.