Squeezing extra performance from existing equipment (sweating the assets') is key to continuous improvement initiatives, particularly during times of cost cutting where the option of spending additional capital to facilitate circuit upgrades can be severely constrained. A common challenge, however, is identifying the best combination of settings for different process variables, as their effects can be camouflaged by interactions with other process variables and, of course, noise in historical data.This paper describes a recent investigation at Boyne Smelters Limited to identify strategies to coarsen the product of its dry, fixed-speed, open-circuit AG mill, without the need for installing pre-screens prior to the mill. The product is used as insulation material for the reduction of alumina to aluminium, and an excessively fine product was inhibiting the ability of the cover material to dissipate heat from the underlying anodes. The approach was to undertake a series of experiments, designed according to statistical criteria, to elucidate the effect of key process variables (feed blend and power draw) upon the product sizing. The results provided clarity about what strategies would encourage the production of a more desirable mill product (with some surprises). The study is a good example of how properly designed experiments can identify hidden potential in processing circuits, thereby mitigating the need for capital-intensive upgrades.CITATION:Vizcarra, T G, Hart, M G, Karbowiak, A T and Napier-Munn, T J, 2018. Getting the most from existing processes: using cleverly designed experiments to optimise the line 3 AG mill at Boyne Smelters Limited, in Proceedings 14th AusIMM Mill Operators' Conference 2018, pp 207-216 (The Australasian Institute of Mining and Metallurgy: Melbourne).