Background: Acute ischemic stroke (AIS) is the fifth leading cause of death in the United States. Although early diagnosis is difficult, no reliable biomarker that categorizes ischemic stroke is currently available. Metabolome profiling is a useful method to identify such biomarkers. Correlation of these biomarkers to objective imaging data such as infarct volume can better delineate stroke pathophysiology and prognosis. Hypothesis: Disturbances of serum metabolites in AIS correlate with infarct volume. Methods: We conducted a prospective pilot study of 20 AIS patients with serum collection in the acute (72 hours) and chronic (2 to 9 months) phases. Global metabolite profiling of each serum sample was performed using Thermo Scientific Q Exactive Plus, and targeted analysis was performed using the in-house metabolomic database. We calculated the volume of ischemic lesions on diffusion-weighted imaging using the ABC/2 formula and correlated it to metabolite levels in the acute phase. The chronic phase served as a reference. We employed the LASSO (least absolute shrinkage and selection operator) regression with hypothesis testing to examine the association of the relative change of metabolite levels in acute and chronic phases with the volume of stroke in the acute phase. Results: The following four metabolites related to oxidative stress showed a significant correlation with infarct volume: 4-pyridoxic acid (B6 metabolism), asparagine (amino acid), serine (sphingolipid metabolism), and citraconic acid (fatty acid) [p-values for all <0.0001]. The infarct volume positively correlated with relative changes of 4-pyridoxic acid and serine (effect size 3.04 (95% CI [2.73, 3.35]) and 1.14 (95% CI [0.56, 1.73]), respectively), while asparagine and citraconic acid negatively correlated with infarct volume (effect size -2.20 (95% CI [-3.47, -0.93]) and -6.75 (95% CI [-10.34, -6.75]), respectively). Conclusion: Although all discovered metabolites were previously associated with ischemia, this is the first study that demonstrates that objective infarct volume correlates with these metabolite levels, which may have prognostic implications. Larger studies in the future are needed to validate our findings.