Gutka, a form of smokeless tobacco, is widely used in the Indian subcontinent and in other regions of South Asia. Smokeless tobacco exposure is most likely to increase the incidence of oral cancer in the Indian population, and metabolic changes are a hallmark of cancer. The development of biomarkers for early detection and better prevention measures for smokeless tobacco users at risk of oral cancer can be aided by studying urinary metabolomics and offering insight into altered metabolic profiles. This study aimed to investigate urine metabolic alterations among smokeless tobacco users using targeted LC-ESI-MS/MS metabolomics approaches to better understand the effects of smokeless tobacco on human metabolism. Smokeless tobacco users' specific urinary metabolomics signatures were extracted using univariate, multivariate analysis and machine learning methods. Statistical analysis identified 30 urine metabolites significantly associated with metabolomic alterations in humans who chew smokeless tobacco. Receiver operator characteristic (ROC) curve analysis evidenced the 5 most discriminatory metabolites from each approach that could differentiate between smokeless tobacco users and controls with higher sensitivity and specificity. An analysis of multiple-metabolite machine learning models and single-metabolite ROC curves revealed discriminatory metabolites capable of distinguishing smokeless tobacco users from nonusers more effectively with higher sensitivity and specificity. Furthermore, metabolic pathway analysis depicted several dysregulated pathways in smokeless tobacco users, including arginine biosynthesis, beta-alanine metabolism, TCA cycle, etc. This study devised a novel strategy to identify exposure biomarkers among smokeless tobacco users by combining metabolomics and machine learning algorithms.