Tiwest employs several methods to assess the quality of heavy mineral concentrates in ore and process samples. Mineralogy is determined using a method called MA98, developed in conjunction with JKMRC over the last three years (Zhang and Whiten, 2001). A range of other techniques are also used to characterise ore, heavy mineral concentrates and mineral products. MA98 is the culmination of efforts since 1989 to determine mineralogy from chemical analysis without recourse to heavy liquid fractionation, which has traditionally been used to discriminate between light-heavy , zircon, rutile and leucoxene. The major advance with MA98 is the use of statistical mass balancing for all chemical species, as opposed to reliance on key chemical species for each mineral. Three magnetic fractions are analysed by X-ray fluorescence spectrometry, with a minerals balance performed for each. Heavy mineral concentrates processed by this method include drill samples, process samples from the wet concentrators and Dry Mill, and mineral products. TBE can be avoided for higher heavy mineral grade samples since quartz is determined from SiO2 balance. In this sense the method is sufficiently universal and robust to cope with a wide range of sample types and compositions. The MA98 algorithm provides abundant information in addition to mineralogy, including TiO2 and Fe2O3 grades of ilmenite and leucoxene, and concentrations of weakly magnetic zircon and kyanite. Information such as this is invaluable in tracking different minerals through the processing plants. Other ore assessment procedures used by Tiwest provide for modelling of orebodies by mineral sizings, Fe in zircon, and FeO and U+Th in ilmenite. Microscopy is used to provide important qualitative information such as grain shape, surface staining and degree of cementing. These quality measurements are included with mineralogy in geological ore models and are available for mine scheduling. In using the above methods to characterise heavy mineral concentrates, Tiwest is better placed to predict and monitor plant performance and product quality from any ore zone.