Ageing presents one of the most fundamental public health challenges of our time. Progress in living standards, combating infectious disease and promoting safety, and general nutritional availability, has led to an increase in lifespan across the developed world. However, this has been accompanied by an increase in the duration of late-life frailty and associated conditions affecting health. Metabolism is known to be a key mediator of ageing across the diversity of living species. Many of the pathways that extend lifespan and promote healthspan are known to be metabolic and relate to managing the balance of energy availability to optimise resource usage and survival during times of scarcity. The model organism Caenorhabditis elegans, a small transparent nematode worm that ordinarily lives in the soil and eats bacteria, is one of the most common organisms used in the study of ageing as it is easy to culture in laboratory conditions and has a short lifespan of around three weeks under normal conditions. In this thesis, I analyse in detail the metabolic changes that occur during ageing in C. elegans, using a multi-omics metabolomics and transcriptomics time series of measurements in three C. elegans strains, and mathematical modelling. Whole-genome metabolic models are representations of all the metabolic reactions taking place within an organism together with their metabolic inputs and outputs, and enzymatic catalysts. I describe the development and validation of a community-wide shared whole-genome metabolic model for C. elegans. Using this model together with measured gene expression levels for each enzyme that catalyses a reaction, it is possible to predict intracellular reaction fluxes using a method called Flux Balance Analysis (FBA). I describe a novel method for the integration of metabolomics data with FBA, and the results of a comparative analysis of the resulting fluxes in normal wild-type ageing. I then go on to describe the differences to a germline-free strain that is long-lived and metabolically different. Finally, I have used the model to probe the metabolic flexibility and evidence for trans-omics bidirectional regulation between the transcriptomic and metabolomic layers.