Metabolomics is a multidisciplinary field with a wide range of applications from basic science to translational medicine. The small molecules it can detect are direct measurements of cellular function and can aptly describe physiological states and/or pathologies even in the absence of phenotypes. The larger the number of identified metabolites, the greater the scope and detail of the metabolic network uncovered. However, metabolomics is limited in the number of metabolites it identifies and their function in a metabolic network. Metabolite identification is technically challenging, with most metabolomics studies relying on chemical standards and metabolite spectral databases that represent only but a small fraction of the metabolites that can be detected (predominantly using mass spectrometry and nuclear magnetic resonance spectroscopy; MS and NMR). Thus, novel strategies are needed for de novo structural elucidation of unknown metabolites. Here, I describe two critical developments to overcome this challenge. First, a method capable of generating ample quantities of a reference material (RM) of any matrix type that is robust to changes over the course of time. Currently there are no easily available metabolomics RMs for the model organism Caenorhabditis elegans. We generated the first metabolomics C. elegans RM. Second, using this RM as a pivotal element, I developed an experimental design that uses semi-preparative fractionation to integrate LC-MS and NMR, two analytical platforms challenging to integrate, and yet essential for the confident identification of metabolites. Metabolomics alone, is necessary but not sufficient to derive mechanistic insight into the interactions of the metabolites it measures with other biomolecules. This functional characterization has been traditionally achieved through biochemical and genetic approaches that strongly rely on model organisms. Using the bacterium Salmonella enterica, I demonstrate that metabolomics can provide a wider view the metabolic effects of 2-amino acrylate (2AA) stress, while carefully planned media supplementation experiments confirmed and expanded on the damage to serine hydroxymethyltransferase and the ensuing effects on the measured metabolites. The complementarity of these three approaches helps addressing two longstanding challenges in metabolomics: metabolite identification and determining their role in the organism; ultimately expanding the tools needed to tackle the complexity of metabolism.