Settling behaviour of polymer conditioned water-treatment sludge was investigated in this study for the purpose of a better understanding of a so-called "CML30 method", which was developed in previous study to evaluate the optimum polymer dosage in sludge conditioning. The "CML30 method" is on the basis of a 30 min settling test in 100 ml measuring cylinders. In this study, the series of settling tests in 100, 500 and 1000 ml measuring cylinders were, respectively, performed and the settling curves (interfacial height to dose and time) were the main focuses for presentation and analyses in great detail in this paper. According to the experimental data presented in this study, settling behaviour may be controlled by both the growth of large sized floc and progressively decreased viscosity. The "speed-up" phenomenon in small polymer dose range obviously enhances the settlement and plays a critical role for the settling set in a 100 ml measuring cylinder at a special dose and at the settling time interval 5-30 min. It is such special observation that leads to the success of so-called ad hoc "CML30 method". However, sludge settling behaviour could be controlled by the formation of networked structure which is involved in excess polymer during the large range (say over 10 mg/l for the case tested) of polymer doses. The higher liquid viscosity values derived from excess polymer will increase the drag force for the resistance of the settlement: In addition, wall effects are likely to interplay with the internal networked structure in large dosed region. (C) 2003 Elsevier B.V. All rights reserved.