A dynamic mathematical model has been developed to predict the performance of a stirred tank, solid-liquid two-phase partitioning bioreactor (SL-TPPB) for the treatment of benzene, toluene, ethylbenzene and o-xylene (BTEX) contaminated gases. The SL-TPPB system consists of an aqueous phase containing a bacterial consortium and a solid phase of silicone rubber beads (10%; v/v)witha high-affinity for BTEX compounds. The silicone rubber beads serve to sequester and release BTEX according to thermodynamic equilibrium, which increases mass transfer from the gas phase and reduces aqueous phase concentrations of these toxic compounds during fluctuating inlet loadings. The model was developed from mass balances on BTEX components in the gas, aqueous and polymer phases, and biomass in the aqueous phase. Dynamic experimental data from this system were used to fit model parameters and to assess the accuracy of the model. A detailed estimability analysis of model parameters and initial conditions was completed to identify uncertain parameters that are most influential for the model predictions and to determine the parameters and initial conditions that should be targeted for estimation using the dynamic data. It was found that the developed model, with estimated parameters and initial conditions, has the ability to predict experimental off-gas BTEX concentrations with reasonable accuracy, which are the outputs of greatest importance. (C) 2009 Elsevier B.V. All rights reserved.