Thickener rakes are essential in the transport of sediment bed material to the underflow in conventional thickeners, however very few studies of bed transport have been published. In this paper, results from pilot-scale thickener experiments with tailor-made yield stress slurries are presented and compared to companion Computational Fluid Dynamics (CFD) simulations. Rake torque is a key issue in thickener operation and it was found that the yield stress of the suspension is the major factor in determining rake torque. Over a range of rake speeds, the measured torque was an almost linear function of yield stress. CFD simulations of the experiments allowed torque to be estimated, and results are shown to be within 20% of the measured values in all cases except the lowest (zero) yield stress suspension. Residence time distributions of solids in the bed were also measured and unusual results were found in which the relationship between residence time and distance from the underflow is not linear (or even monotonic). CFD results clearly show that for uniform sized rake blades, the over-delivery of an outer blade (compared to the next inner blade) sets up recirculation in the bed, especially in the outer regions of the tank, and this can result in long material pathways and hence long residence times. This picture is further complicated by the relative contributions of rake delivery and underflow rate, and indicates that a simple picture of plug flow in the bed is far from reality. The study illustrates the value that can be obtained from validated CFD modelling of thickener rakes. Crown Copyright (C) 2008 Published by Elsevier B.V. All rights reserved.