Two stoichiometric black-box models were compared in glucose and ethanol batch cultures of Monascus ruber to estimate some bioprocess variables. Although mass balance models are usually applied to a steady-state process, their adaptation to batch processes provided a good correlation between experimental and predicted values of biomass and monosodium glutamate for two different ethanol cultures. In contrast, the empirical stoichiometric model requires more analytical information and the estimation of variables and poores correlations were obtained for the joint estimation of biomass production and monosodium glutamate consumption Close correlation between experimental and predicted data was found for the joint estimation of biomass and malate with this model.