Improved understanding of the variables that influence hydrocyclone performance is of increasing interest to the iron ore industry, where hydorcyclones are widely used to deslime fine ore. In many cases, the impact of particular variables is not quantified sufficiently well experimentally to ensure that industrial performance will match expectations. The current work focused on developing and implementing a systematic methodology to evaluate and compare the performance of four different hydrocyclones in a desliming application. A critical component of the project was the design and construction of a pilot-scale hydrocyclone rig using state of the art monitoring and sampling methods. A wide ranging test programme was conducted to quantify the effect on metallurgical performance of the key hydrocyclone variables such as pressure, feed density, vortex finder size, spigot size and hydrocyclone make/overall design. Details of the experimental approach are presented, along with a discussion of the sizing analysis methods used. Performance criteria were established to assess the four hydrocyclones. The effects of each major variable were quantified over a wide range of conditions. The results show that clear performance differentiation was achieved, enabling the selection of the optimum configuration. The results from this work demonstrate the benefit of objectively assessing the performance of nominally equivalent units in selecting an optimum configuration that best satisfies the performance criteria.