<p>Plant metabolomics has become a powerful approach for unravelling the chemical complexity of plants and their responses to genetic and environmental factors. While significant progress has been made with targeted and untargeted strategies using LC–MS, GC–MS, IMS, ULC–IMS–QTOF, FT-ICR-MS, and NMR, the field still faces major challenges. Comprehensive metabolomics coverage remains elusive due to vast chemical diversity, metabolite instability, and incomplete reference libraries, which hinder accurate identification and biological interpretation. Furthermore, integrating outputs from diverse analytical platforms into meaningful biological insight requires robust computational tools and standardized workflows, yet gaps persist in data harmonization, reproducibility, and species-specific databases. This review addresses these critical challenges, synthesizing current advances in plant metabolomics, highlighting the complementary role of targeted and untargeted approaches, and underscoring the emerging contribution of NMR to structural elucidation and dynamics pathway analysis. More importantly, it emphasizes the need for improved databases and scalable bioinformatics and integration with multi-omics and machine learning to fully exploit metabolomics for crop improvement, sustainable agriculture, and transitional research. By outlining both technological progress and unresolved limitations, this review provides a timely roadmap for advancing plant metabolomics toward more comprehensive and impactful applications.</p> Graphical abstract <p></p>