The central nervous system (CNS) serves as the core for regulating human body functions. The structural and metabolic networks of brain regions exhibit significant heterogeneity, with distinct brain regions having precisely divided metabolic characteristics. For CNS diseases such as Alzheimer's disease (AD), Parkinson's disease (PD), and ischemic stroke (IS), their pathological mechanisms are often accompanied by metabolic disorders in specific brain regions. Notably, such metabolic perturbations precede macroscopic structural damage, thus providing a promising avenue for the identification of biomarkers for the early diagnosis of these disorders. However, traditional diagnostic methods and therapeutic effect assessment approaches have limitations. Specifically, they fall short in identifying subtle abnormalities in early metabolic levels, characterizing the spatial distribution of metabolites, and accurately determining whether treatments can reverse the metabolism-related pathological disorders. Based on mass spectrometry imaging technology, spatial metabolomics enables in situ detection and spatial localization of hundreds of metabolites in CNS tissue sections at a spatial resolution of the micrometer level. It not only fills the gap in technologies for detecting early metabolic abnormalities in diseases but also drives the research on disease pathological mechanisms from the "macroscopic structure" to the "molecular spatial network" level. Consequently, it provides crucial technical support for the screening of disease biomarkers, the analysis of disease mechanisms, and the evaluation of drug therapeutic effects. This article systematically reviews the main technologies of spatial metabolomics and deeply explores its current application status in CNS disease research. The aim of this review is to offer theoretical support and practical references for advancing the development of metabolomics in this field.