The idea behind Grade Engineering is to exploit the natural heterogeneity of an orebody (spatial, temporal) in order to divert low-value waste out of the mining value chain. Essentially, this involves rejecting material at relatively coarse particle sizes (10 mm to 200 mm). However, lack of knowledge about the inherent particle grade distributions in these coarse particle size ranges limits the opportunities for optimisation. Particle grade distributions are essential for quantifying and modelling gangue liberation and yield curves. These curves can then be used to predict the mass and value flows that will go towards the accept and the reject streams. Particle grade distribution information for particles coarser than 10 mm can be collected based on chemical assays of sufficient number of individual particles. However, for smaller particles (1 mm to 9 mm), a significantly larger number of particles is required for assays. Moreover, in some cases, the mass of individual particles is smaller than the minimum mass required for chemical assaying. Also, other quantification techniques such as automated mineralogy, eg Mineral Liberation Analysis (MLA), require extended measurement times to analyse statistically representative numbers of particles at these size ranges. This work investigates the use of 3D X-ray microtomography to fill this quantification gap at these size ranges (1 mm to 9 mm). In-house image processing algorithms are used to estimate the mineral/metal content in each particle. Particle grade distributions, as well as density distributions, are thus produced without a minimum mass requirement and with each scan providing information about tens to hundreds of particles at a time. Mineralogical data from MLA analysis, of finer size ranges, is used to inform X-ray microtomography, so a simpler 3D mineral map can be obtained using the image intensity. Chemical assay information of the scanned samples is used to validate and correct the mineral phase segmentation.