A cascade sequence of four biological aerated filters was used to examine the treatment of a simulated textile waste. The dye component of the synthetic wastewater was Maxilon Red BL-N, which was added at 25 and 50 mg/litre. The presence of the dye caused a very slight deterioration in the performance of the overall system in terms of COD removal. A high colour removal (99%) was achieved with both the concentrations tested, with the first reactor removing the majority (96%). Adsorption studies showed that the dye-sludge system conformed to the Langmuir equation but the application of a mass balance model for a multi-stage reactor system showed discrepancies between the calculated dye concentrations and those measured experimentally. It is suggested that this was due to competition for the adsorption sites. A comparison of the binding potential of the reactors (based on the Langmuir maximum adsorption capacity) with the mass of dye removed during the test periods suggests that a significant amount of biodegradation is taking place. (C) 1997 Elsevier Science Ltd.