Osteoarthritis (OA) is a multifactorial and progressive disease affecting whole synovial joint. The extract pathogenic mechanisms and diagnostic biomarkers of OA remain unclear. In this article, we review the studies related to metabolomics of OA, discuss the biomarkers as a tool for early OA diagnosis. Furthermore, we examine the major studies on the application of metabolomics methodology in the complex context of OA and create a bridge from findings in basic science to their clinical utility. Recently, the tissue metabolomics signature permits a view into transitional phases between the healthy and OA joint. Both nuclear magnetic resonance spectroscopy (NMR) and mass spectrometry-based metabolomics approaches have been used to interrogate the metabolic alterations that may indicate the complex progression of OA. Specifically, studies on alterations pertaining to lipids, glucose, and amino acid metabolism have aided in the understanding of the complex pathogenesis of OA. The discovery of identified metabolites could be important for diagnosis and staging of OA, as well as for the assessment of efficacy of new drugs.