Predicting the physical characteristics of various ore types has for long been a difficult problem during mining. Grade control practices tend to concentrate on the chemical grade of the ore. Estimation of physical characteristics such as beneficiation behaviour is commonly limited to the simplistic application of historical averages. Large amounts of geochemical data are often gathered during the exploration and mining stages of a project, especially in ferrous and base metal deposits. In addition to the elements of principal economic interest, samples are often analysed for suites of gangue constituents to determine whether the mineralisation lies within target specifications. This multi-element data is often put to no other use than this simple chemical classification, yet it may present the geologist or metallurgist with valuable insights into the mineral content and physical characteristics of the deposit. Normative mineral estimates have been used as a method for characterising igneous rocks for almost 100 years. The method establishes and applies a set of sequential rules, by which the chemical components of a sample are assigned to mineral components. Provided that it is possible to make some realistic assumptions about the mineral species likely to be present in a rock unit or deposit and their paragenesis, this approach can be extended to a variety of mineral deposits. The geochemical data that are routinely collected when, for example, drilling iron ore and base metal deposits, can be used to derive estimates of normative mineral content. These estimates can be used to characterise physical or metallurgical aspects of the ore, such as density, beneficiation behaviour and metal recovery. This paper examines the application of normative mineral estimates to a variety of practical resource characterisation problems. The potential for predicting the lump and fines proportions of iron and manganese ores is also discussed.