There is little published literature on the design and operation of rakes in gravity thickeners. In practice, design principles and operating guidelines rely heavily on experience and are usually qualitative in nature. In some industry applications, in particular in minerals processing, raking can be a rate-limiting step, with high rake torque, rake "bogging", rat-holing and insufficient underflow density all attributed to poor rake design or operation. In this paper we define and discuss a range of efficiency measures that allow different aspects of raking in gravity thickeners to be quantified. We undertake Computational Fluid Dynamics (CFD) simulations in a model of the settled bed in a 45 m diameter thickener and apply these measures to the simulation results in order to illustrate the information that they provide. The measures suggest that the ratio of rake delivery and underflow rate plays a key role in designing a rake system. Other important factors to consider are the relative delivery of neighbouring rake blades, drag forces that determine rake torque and the degree of "swirl" motion imparted to the bed. All of these factors can be taken into consideration to modify and improve rake design and operation. Rake speed affects delivery to the underflow but does not play a significant role in other performance measures so can (in principle) be adjusted to balance the desired underflow rate and rake delivery without affecting performance. The results of this paper suggests that the application of a validated CFD raking model coupled to the use of raking efficiency measures is a good starting point for the quantitative assessment of design and operation of thickener rakes. The results also suggest that pilot-scale raking studies are a valid approach when investigating raking on the full-scale. Crown Copyright (C) 2010 Published by Elsevier B.V. All rights reserved.