Introduction: Cerebral atrophy after stroke is associated with poor functional outcome. The prediction and prevention of post-stroke brain atrophy could therefore represent a target for neurorestorative therapies. We investigated the associations between peri-infarct metabolite concentrations measured by quantitative MRS and brain volume change in the infarct hemisphere after stroke. Methods: Twenty patients with ischemic stroke were enrolled. Patients underwent 3T-MRI within 1 week of onset, and at 1 and 3 months. At the baseline scan, an MRS voxel was placed manually in the peri-infarct area and another in the corresponding contralateral region. Volumetric analysis of T1 images was performed using two automated processing packages. Changes in gray and white matter volume were assessed as percentage change between 1 and 3 months. Results: Mean concentrations (institutional units) of N-acetylaspartic acid (NAA) (6.1 vs 7.0, p = 0.039), total creatine (Cr+PCr) (5.4 vs 5.8, p = 0.043), and inositol (4.5 vs 5.0, p = 0.014), were significantly lower in the peri-infarct region compared with the contralateral hemisphere. There was a significant correlation between baseline peri-infarct NAA and white matter volume change in the infarct hemisphere between 1 and 3 months, with lower NAA being associated with subsequent white matter atrophy (Spearman's rho = 0.66, p = 0.010). The baseline concentration of Cr+PCr was also significantly correlated with white matter atrophy in the infarct hemisphere (Spearman's rho = 0.59, p = 0.027). Both of these associations were significant after adjustment for the false discovery rate and were validated using the secondary volumetric method. Conclusion: MRS may be useful in the prediction of white matter atrophy post-stroke and in the testing of novel neurorestorative therapies.