A series of flotation experiments was carried out in three Rushton turbine cells of volumes 2.25, 10 and 50 dm(3) using hydrophobic quartz particles to determine a set of scale-up criteria that would produce the same size-by-size flotation rate constants. Flotation was performed at a constant Sauter mean bubble diameter over a range of superficial gas velocities and impeller rotational speeds. The overall flotation rate constant increased linearly with increasing superficial gas velocity (and hence bubble surface area flux). The rate constant also increased linearly with increasing energy dissipation, until a maximum value was reached. A further increase in energy dissipation had little effect on the rate constant. The dependency of the rate constant on energy dissipation is a reflection of the size range and hydrophobicity of the particles used in this study. Good agreement was obtained between the experimental results and those predicted by a fundamental flotation model using experimentally measured values for mean energy dissipation and the Sauter mean bubble diameter. The bubble velocity was adjusted to obtain the best fit of the experimental data. This inferred bubble velocity was found to increase with increasing superficial gas velocity and decrease with increasing impeller rotational speed. To achieve scale-up of the flotation rate constants, maintaining a constant bubble surface area flux was used as one criterion. The other operating variable was the impeller rotational speed, which, for successful scale-up was adjusted so that N(3)D was constant over the cell volume range. This enabled the measured mean energy dissipation for the three cells to be held constant as the cell volume increased. (c) 2006 Elsevier B.V. All rights reserved.