Abstract: Single-cell metabolomics enables real-time dynamic tracking of small-molecule metabolites (with molecular weights ranging from 50 to 2 000), including amino acids, nucleotides, lipids, and their derivatives, within individual cells. However, it confronts distinctive challenges. Unlike nucleic acids or proteins, metabolites cannot be amplified, thus demanding ultra-high sensitivity from analytical tools. They exhibit enormous variations in chemical properties, ranging from highly polar hydrophilic molecules to non-polar lipids, and span an extremely wide abundance range, posing rigorous requirements for the universality of analytical methods. Additionally, the existence of numerous isomers (such as lipid C＝C isomers) further complicates accurate identification, making comprehensive and precise analysis an arduous task. Traditionally, single-cell metabolomics predominantly relied on direct ionization mass spectrometry (MS) techniques, including electrospray ionization (ESI), matrix-assisted laser desorption/ionization (MALDI), and secondary ion mass spectrometry (SIMS). Alternatively, it combined these ionization methods with micro-separation techniques such as capillary electrophoresis (CE) and ion mobility spectrometry (IMS) to compensate for the insufficient structural resolution of direct ionization approaches. While these methods have yielded substantial data, they are constrained by limited structural identification capabilities. Moreover, unseparated direct ionization suffers from severe signal masking, where high-abundance metabolites (such as phospholipids) often overshadow low-abundance but biologically significant ones, significantly restricting analytical coverage. Over the past five years, with optimized single-cell metabolite extraction and pretreatment protocols, coupled with advancements in instrumentation such as nano-liquid chromatography (nanoLC) and high-resolution mass spectrometry (HRMS), liquid chromatography-mass spectrometry (LC-MS) has gradually emerged as a pivotal approach in single-cell metabolomics, marking a paradigm shift toward standardization. NanoLC, characterized by its low flow rate, minimizes sample dilution, thereby enhancing sensitivity. Chemical derivatization techniques have also made notable contributions by introducing functional groups to phosphorylated metabolites. In lipidomics, direct ionization methods (ESI, MALDI, SIMS) have generated extensive lipid profiles, but their inability to accurately distinguish isomers limits their application. LC-MS, particularly when coupled with ion mobility spectrometry (IMS), has effectively addressed this issue. Single-cell multi-omics analysis, which integrates metabolomics with proteomics and transcriptomics, has also made significant strides by leveraging platforms such as tandem cytometry. Looking ahead, the field of single-cell metabolomics will focus on several key directions: developing online pretreatment technologies such as supercritical fluid extraction (SFE) to improve extraction efficiency and reduce manual intervention, constructing highly integrated automated microfluidic platforms to enhance throughput, and deeply integrating artificial intelligence algorithms for data mining to standardize data processing workflows. These advancements are expected to further boost sensitivity, throughput, and analytical coverage, propelling single-cell metabolomics toward broader applications in precision medicine, such as guiding personalized cancer therapy and facilitating early disease diagnosis.