A wealth of quantitative data on the types, volume distribution, and mode of occurrence of the platinum-group minerals residing in the Merensky Reef, the UG-2 chromitite layer, and Platreef has been obtained over a twelve-year period with the result that general platinum-group mineral types for unexplored areas can perhaps now be predicted. The platinum-group minerals of the UG-2 and overlying Merensky layers, although locally different in composition, mode of occurrence, and size, show close ties when viewed on a regional scale, especially with regard to their Pt-Pd sulfide and Pt-Fe alloy contents. Thus, if in one particular area the Merensky Reef contains Pt-Pd sulfides as the predominant platinum-group minerals, then it can be safely inferred that Pt-Pd sulfides will be a significant component of the UG-2 layer in that area. The occurrence of Pt-Fe alloy also follows this principle. The distribution of Pt-Fe alloy is in fact a major factor in differentiating the ores in the vicinity (north and south) of the Pilansberg Alkaline Complex from those of the same layers farther along the arcuate strike (northeastern and southeastern horns of the western Bushveld) and has led to an overall concept relating platinum-group mineral types to postulated centers of increased volatile activity or perhaps feeder sites of the complex. The mineralogy of reef disturbances such as potholes is discussed. The mode of occurrence of the platinum-group minerals varies between base metal sulfide, silicate, and chromite associations and the size of platinum-group minerals in the various reefs is given. Solid solution or lattice substitutions of platinum-group elements in the base metal sulfides account for the very important latent fraction of the total platinum-group elements (as distinct from platinum-group minerals) of the Bushveld Complex. Rhodium in the Merensky Reef occurs almost wholly in solid solution within pentlandite, whereas Rh minerals occur in the UG-2 layer. Chromite tends to host microscopic to submicroscopic inclusions of laurite and does not appear to take Ru into the crystal lattice. The probability of solid solution of platinum-group elements in silicate and other oxide phases is discussed.