The metabolic fluxes through the central carbon pathways were calculated for the genus Bacillus separately for the enzymes serine alkaline protease (SAP), neutral protease (NP) and alpha -amylase (AMY) on five carbon sources that have different reduction degrees (gamma), to determine the theoretical ultimate limits of the production capacities of Bacillus species and to predict the selective substrate for the media design. Glucose (gamma = 4.0), acetate (gamma = 4.0), and the TCA cycle organic-acids succinate (gamma = 3.5), malate (gamma = 3.0), and citrate (gamma = 3.0) were selected for the theoretical analyses and comparisons. A detailed mass flux balance-based general stoichiometric model based on the proposed metabolic reaction network starting with the alternative five carbon sources for the synthesis of each enzyme in Bacillus licheniformis that simulates the behaviour of the metabolic pathways with 107 metabolites and 150 reaction fluxes is developed. Highest and lowest specific cell growth rates (mu) were calculated as 1.142 and 0.766 h(-1), respectively, when glucose that has the highest degree of reduction and citrate that has the lowest degree of reduction were used as the carbon sources. Highest and lowest SAP, NP and AMY synthesis rates were also obtained, respectively, when glucose and citrate were used. Metabolic capacity analyses showed that the maximum SAP, NP, and AMY synthesis rates were, respectively, 0.0483, 0.0215 and 0.0191 mmol g(-1) DW h(-1) when glucose uptake rate was 10 mmol g(-1) DW h(-1) and specific growth rate was zero. The amino acid compositions and the molecular weights of the enzyme influence the production yield and selectivity. For SAP and NP oxaloacetate and pyruvate, for AMY oxaloacetate appear to be the critical main branch points. Consequently, for SAP and NP syntheses the fluxes towards the alanine group and aspartate group, and for AMY synthesis the flux towards the aspartate group amino acids need to be high. The results encourage the discussion of the potential strategies for improving productions of SAP, NP and AMY. (C) 2001 Elsevier Science B.V. All rights reserved.