Metabolic network reconstruction has become an established procedure that allows the integration of different data types and provides a framework to analyze and map high-throughput data, such as gene expression, metabolomics,and fluxomics data. In this chapter, we discuss how to reconstruct a metabolic network starting from a genome annotation. Further experimental data, such as biochemical and physiological data, are incorporated into the reconstruction, leading to a comprehensive, accurate representation of the reconstructed organism, cell, or organelle. Furthermore, we introduce the philosophy of constraint-based modeling, which can be used to investigate network properties and metabolic capabilities of the reconstructed system. Finally, we present two recent studies that combine in silico analysis of an Eschirichia coli metabolic reconstruction with experimental data. While the first study leads to novel insight into E. coli's metabolic and regulatory networks, the second presents a computational approach to metabolic engineering.