The metabolic network of Xanthomonas campestris is complex since a number of cyclic pathways are present making simple stoichiometric yield predictions difficult. The influence of certain pathway configurations and the resulting variations in flux have been examined as regards the maximum yield potential of this bacteria for xanthan gum production. These predictions have been compared with experimental results showing that the strain employed is functioning close to its theoretical maximum as regards yield criteria. The major constraint imposed on the network concerns energy availability which has a more pronounced effect on yield than carbon precursor supply. This can be attributed to the relatively high maintenance requirements determined experimentally and incorporated into the model. While some of this overall energy burden will undoubtedly be associated with incompressible metabolic requirements such as sugar uptake and xanthan efflux mechanisms, future strain improvement strategies will need to attack other non-essential energy-consuming reactions, if yields are to be further increased.