In this article, we present yet another application of the Attainable Region (AR) method to data from a laboratory scale milling of a low grade gold ore. In this particular case, we investigate how to optimize the amount of material in a desired size class for a scenario where the boundaries of the desired size class of interest are changed. The AR approach has never been applied in such a scenario before. Using a mono sized feed of -1700 + 850 mu m, two desired product size classes of interest (-850 + 150 mu m and -150 + 75 mu m) are selected in order to determine the optimum interstitial filling (U) to be used in the mill. Two different values of U, 1.75 and 1.0 respectively, are obtained as, optimal. Our results also show that optimal operating conditions are different for different objective functions. We demonstrate that the AR may be used to specify optimal conditions that may be used for particle size reduction processes. We also illustrate how an AR boundary for optimization purposes can in certain cases be extended using the 'mixing principle'. (C) 2017 Elsevier B.V. All rights reserved.