Updated project metadata. Background & aims: Stroke diagnosis is challenging in the acute phase. We aimed to determine biomarkers for the differentiation of ischemic stroke (IS) from intracerebral hemorrhage (ICH) using SWATH-MS and validate the discovered biomarkers within 24 hours using MRM proteomics. Methods: The study was conducted at Department of Neurology, All India Institute of Medical Sciences, New Delhi, India. Serum samples were collected within 24 hours from acute stroke (IS & ICH) patients & healthy controls. SWATH-MS proteomics identified significantly differentially expressed (SDE) proteins (fold change: 1.5, p<0.05 and confirmed/tentative selection using Boruta random forest) between IS and ICH which were then validated using protein MRM. Cut-off points were determined using Youden Index. Prediction models were developed using forward stepwise multivariate logistic regression analysis. Integrated discrimination improvement index determined the added value of biomarkers to clinical models. Results: Discovery phase included 20 IS, 20 ICH and 40 controls while validation phase included 150 IS and 150 ICH subjects. Total 365 proteins were quantified using SWATH-MS. Between IS and ICH, 20 SDE proteins were identified. Prediction model including clinical variables (sex, hypertension, atrial fibrillation, current smoking, baseline NIHSS) and biomarkers (GFAP, MMP9, APOC1) independently differentiated IS from ICH (accuracy: 92%, sensitivity: 96%, specificity: 69%). Addition of biomarkers improved the discrimination capacity by 26% (p<0.001) compared to clinical variables alone. Conclusion: Our study identified a range of potential protein biomarkers for the differentiation of IS. Protein biomarkers along with clinical predictors might be useful candidates in differentiating IS from ICH in acute settings.