Batch tests were employed to estimate the optimal conditions for excess sludge reduction under an alternating aerobic/oxygen-limited environment using response surface methodology. Three key operating parameters, initial mixed liquor suspended solids (initial MLSS), HRT (hydraulic retention time) and reaction temperature (T), were selected, and their interrelationships studied by the Box-Behnken design. The experimental data and ANOVA analysis showed that the coefficient of determination (R-2) was 0.9956 and the adjR(2) was 0.9912, which demonstrates that the modified model was significant. The optimum conditions were predicted to give a maximal Delta MLSS yield of 226 mg/L at an initial MLSS of 10,021 +/- 50 mg/L, an HRT of 9.1 h and a reaction temperature of 29 degrees C. The prediction was tested by triplicate experiments, where a Delta MLSS yield of 233 mg/L was achieved under the chosen optimal conditions. This excellent correlation between the predicted and measured values provides confidence in the model. (C) 2011 Elsevier Ltd. All rights reserved.