Abstract Background Metabolomics is increasingly recognized for uncovering pathophysiology and identifying diagnostic and prognostic biomarkers in inflammatory bowel disease (IBD). This study focuses on exploring comprehensive serum metabolomic profiles and corresponding specific metabolic pathways, in order to differentiate IBD patients from healthy individuals and to accurately identify IBD subtypes (Crohn's disease [CD], ulcerative colitis [UC]) through serum biomarkers with high patient acceptability. Methods Serum samples with matched clinical metadata were comprehensively collected from patients with IBD at Kyung Hee University Hospital, Seoul, Korea. Corresponding samples from normal controls (NCs) were obtained for comparison. Untargeted profiling was performed with Gas Chromatography-Time-Of-Flight-Mass Spectrometry, and targeted profiling of bile acids and tryptophan used Liquid Chromatography-Triple Quadrupole-Mass Spectrometry. The identification of distinct metabolites and potential biomarkers was achieved through extensive univariate and multivariate statistical analyses. Receiver operating characteristic (ROC) curves and metabolic pathway analyses were also conducted. Results A total of 346 subjects were analyzed, comprising 258 IBD patients (134 with CD and 124 with UC) and 88 NCs, Table 1. The patients with IBD were clearly clustered from the NCs, while IBD subgroups were not clearly separated from each other in the nontargeted profiling. Tryptophan and indole-3-acetic acid were elevated in both patients with CD and UC, while kynurenine and indole-3-propionic acid increased only in CD. Conversely, patients with UC had lower levels of indole-3-acetic acid, serotonin, and acetylcholine compared to CD. The ratio of primary and secondary bile acids was also significantly decreased in both patient groups compare with NCs. The ROC curves (Figure 1A) showed good discriminatory power for NCs vs CD (AUC:0.9738), NCs vs UC (AUC:0.9887), and UC vs CD (AUC:0.7140). Pathway analysis of the identified metabolites showed a sustained change in glyoxylate and dicarboxylate metabolism/alanine, aspartate and glutamate metabolism/glycine, serine and threonine metabolism associated with all comparisons (NCs vs UC, NCs vs CD, UC vs CD). Beta-alanine, arginine, and proline metabolism were linked to IBD compared to NCs. Glycerolipid metabolism distinctly differed between UC and CD (Figure 1B-1D). Network analysis revealed associations between metabolomic markers and specific clinical phenotypes of IBD subtypes. Conclusion This study demonstrates that serum metabolomic biomarkers are novel and effective tools for diagnosing IBD, as well as for identifying and characterizing its subtypes and associated phenotypes.