The organic carbon in highly variable dye wastewater could be decomposed by microorganisms. The carbon turnover had some similarity to the transformation of soil organic matter, in that both systems had complicated carbons in the initial, intermediate and final phases, with microorganisms activating the decomposition. These processes could be simulated using a dynamic model that includes a system of differential equations representing six state variables, namely the liable decomposable fraction (LDF), the resistant decomposable fraction (RDF), active biomass (BAC), inactive biomass (BIA), decomposable (DOM) and non-decomposable organic materials (NDM). The relationship between these variables and traditional wastewater properties, e.g. chemical oxygen demand (COD) and biological oxygen demand (BOD), was defined in the context of experimental data acquisition and model evaluation. The model was calibrated by simulating several conditions, such as increasing the organic loading and hydraulic retention time (HRT), varying the C/N ratio of the organic substances in the dye wastewater and changing the treatment temperature. With the data from these experiments, a set of parameters related to the transformation rate and proportion was identified to represent the results from each experiment. This fitting allowed the model to be useful in carrying out more simulations for characterizing the efficiency of different treatment strategies. The results showed that in order to meet discharge COD regulations, <94 mg C/l in the effluent, the extension of HRT was more effective and sensitive than reducing the organic loading. Both the simulations and experiments indicated that the organic loading limit for the system was near 2400 mg/l COD. The simulation also revealed that the extension of HRT significantly improved the RDF and NDM transformations. No significant changes were found for BAC or BIA as the C/N ratio increased, resulting in minor effects on the other organic substance variations. The temperature factor affected the microbial activity, and therefore the entire transformation process. The effects were typically a parabolic function with the optimum between 20 and 30 °C. When the temperature was increased from 40 to 50 °C, the effluent COD increased dramatically.