Halomonas smyrnensis AAD(T) is a halophilic, gram-negative bacterium that can efficiently produce levan from sucrose as carbon source via levansucrase activity. However, systems-based approaches are required to further enhance its metabolic performance for industrial application. As an important step toward this goal, the genome-scale metabolic network of Chromohalobacter salexigens DSM3043, which is considered a model organism for halophilic bacteria, has been reconstructed based on its genome annotation, physiological information, and biochemical information. In the present work, the genome-scale metabolic network of C. salexigens was recruited, and refined via integration of the available biochemical, physiological, and phenotypic features of H. smyrnensis AAD6(T) . The generic metabolic model, which comprises 1,393 metabolites and 1,108 reactions, was then systematically analyzed in silico using constraints-based simulations. To elucidate the relationship between levan biosynthesis and other metabolic processes, an enzyme-graph representation of the metabolic network and a graph decomposition technique were employed. Using the concept of control effective fluxes, significant links between several metabolic processes and levan biosynthesis were estimated. The major finding was the elucidation of the stimulatory effect of mannitol on levan biosynthesis, which was further verified experimentally via supplementation of mannitol to the fermentation medium. The optimal concentration of 30 g/L mannitol supplemented to the 50 g/L sucrose-based medium resulted in a twofold increase in levan production in parallel with increased sucrose hydrolysis rate, accumulated extracellular glucose, and decreased fructose uptake rate.