in Wiley Online Library (wileyonlinelibrary.com).The cybernetic approach to metabolic modeling tracing its progress from its early beginnings to its current state with regard to its relationship to other modeling approaches, applications to bioprocess modeling, metabolic engineering, and future prospects are described. The framework is shown to handle large metabolic networks in making dynamic predictions from limited data with looming prospects of extending to genome scale networks. V V C 2012 American Institute of Chemical Engineers AIChE J, 58: 986-997, 2012 Keywords: dynamic metabolic models, the cybernetic approach, metabolic engineering, elementary modes, cellular regulation where W j denotes the species whose concentration is w j . Eq. 1 represents n r reactions of metabolism which subsume both transport and chemical reaction with a ij the stoichiometric coefficient which, as per the usual convention, is positive, negative, or zero, in respective accordance with W j being a product, reactant, or nonparticipant in the ith reaction. The rate associated with the ith reaction is denoted by r i . As each reaction is catalyzed by a specific enzyme, the level of this enzyme and its activity would determine the rate of the reaction besides the concentrations (as dictated by the kinetics) of species participating in the reaction.Predictions (a) based on measurements of glucose alone leads to good predictions of biomass, formate, ethanol but not of lactate and succinate. When measurements of glucose and lactate are both used for model identification (b), all predictions (shown in continuous lines) are considerably improved. Figure 13. Cybernetic model simulations of Namjoshi et al. 52 alongside with experimental data of Europa et al., 50 showing multiple steady states in continuous cultures of hybridoma cells.