Abstract : Lipid quantity and their fatty-acid composition are partly responsible for meat quality traits and play a major role in many pathologies. A better control of the lipid metabolism is then important in terms of public health and food production. Numerous experimental and biblio- graphical data now available on lipid metabolism ; however both the hierarchy of importance of metabolic pathways and the key regulators remain unhnown under various conditions. Mod- elling tools may be helpful to a better understanding of lipid metabolism in various organisms. Two complementary models were developed: a simple dynamic model based on a minimum set of biological functions and regulations necessary to explain experimental data observed on lipid metabolism, and a large-scale model including a maximum of information extracted from knowledge databases. The ﬁrst model is composed of a set of ordinary diﬀerential equations describing the main biochemical pathways involved in lipid metabolism, independently of the species, the organ and experimental conditions. This model was confronted with biological data describing the variation of fatty-acid content in the liver and adipose tissue of two types of mice during a 72 hours fasting : wild-type mice and knockout for PPARα (i.e., a transcription factor that mediates fatty acid oxidation during fasting). Liver fatty acid inﬂux from blood, fatty acid oxidation and unexpected elongation and desaturation of fatty acid were identiﬁed as important during fasting in mice liver. Morever the existence of an unknown activator of elongation and desaturation diﬀerent from PPARα was suggested. The second model consists on a regulatory network which aims at gathering literature knowl- edge, and to compare it to high throughput transcriptomic data. Three literature databases focusing on biological interactions were compared(Gardon, i.e., a small database handmade by experts of lipid metabolism, and TRANSPATH and Ingenuity, i.e., two commercial databases of biological interactions). A strong complementarity between these three databases regarding their sources and the exploitable contents in terms of extractable inﬂuences was shown. A regulatory network was then built by merging their automatically-extracted contents and the most connected elements related to energy metabolism were found by topological analysis. Models are then useful to identify know and unknown regulators and pathways, and suggest new biological experiments.