While much work has concentrated on focal white matter (WM) lesions in multiple sclerosis, there is growing evidence to suggest that normal-appearing WM (NAWM) and grey matter (GM) are also involved in the disease process. This study investigated multiple sclerosis disease effects on NAWM and cortical GM (CGM) metabolite concentrations, and the relationships between these metabolite concentrations and clinical impairment. Proton magnetic resonance spectroscopic imaging ((1)H-MRSI) data acquired using point resolved spectroscopic (PRESS) localization (echo time 30 ms, repetition time 3000 ms, nominal voxel volume 2.3 ml) from 27 relapsing-remitting multiple sclerosis and 29 normal control (NC) subjects were processed using LCModel to estimate metabolite concentrations in millimoles per litre. (1)H-MRSI voxel tissue contents were estimated using SPM99 tissue and semi-automatic lesion segmentations of three-dimensional fast spoiled gradient recall scans acquired during the same scanning session. NAWM and CGM metabolite concentrations estimated were: choline-containing compounds (Cho); creatine and phosphocreatine (Cr); myo-inositol (Ins); N-acetyl-aspartate plus N-acetyl-aspartyl-glutamate (tNAA); and glutamate plus glutamine (Glx). CGM data came from 24 of the multiple sclerosis (mean age 35.2 years, mean disease duration 1.7 years) and 25 of the NC (mean age 34.9 years) subjects. NAWM data came from 25 of the multiple sclerosis (mean age 35.0 years, mean disease duration 1.7 years) and 28 of the NC (mean age 36.7 years) subjects. Metabolite concentrations were compared between multiple sclerosis and NC subjects using multiple (linear) regression models allowing for age, gender, (1)H-MRSI voxel tissue and CSF contents, and brain parenchymal volume. At a significance level of P < 0.05, CGM Cho, CGM and NAWM tNAA, and CGM Glx were all significantly reduced, and NAWM Ins was significantly elevated. Spearman correlations of multiple sclerosis functional composite scores with tissue metabolite concentrations were significant for the following: CGM Cr (r(s) = 0.524, P = 0.009), CGM Glx (r(s) = 0.580, P = 0.003) and NAWM Ins (r(s) = -0.559, P = 0.004). These results indicate that metabolite changes in NAWM and CGM can be detected early in the clinical course of multiple sclerosis, and that some of these changes relate to clinical status. The correlation of clinical impairment with CGM Cr and Glx but not tNAA suggests that it is more closely associated with neuronal metabolic dysfunction rather than loss in clinically early relapsing-remitting multiple sclerosis. The correlation of clinical impairment with a raised NAWM Ins may indicate that glial proliferation also relates to function at this stage of the disease.