Platiniferous horizons such as the Merensky Reef that are mined chiefly for their PGE contents constitute one end member of a continuum of Ni-Cu-PGE sulfide deposits; Ni-(Cu) sulfide deposits in which the PGE may or may not be by-products represent the other end of this continuum. Although it has long been thought that PGE-rich sulfides can only be developed by magma mixing in large magma chambers, new PGE data for the Bushveld Complex suggests an alternate model to account for origin of mineralization in the world's most important Pt producer. Briefly, this model suggests that PGE-rich sulfides were initially deposited at depth below the Bushveld Complex as a result of interaction of a S-undersaturated mantle magma with S-rich crustal rocks. These PGE-rich sulfides were subsequently re-dissolved by a second magma, which introduced the PGE into the Bushveld Complex magma chamber, where the PGE-rich reefs were formed due to cooling of these magmas. If this model is correct, it means that large layered mafic/ultramafic intrusions are not a prerequisite for the formation of economic PGE-rich deposits. Exploration for Merensky Reef-type PGE deposits can be greatly facilitated by documenting variations in the background PGE contents of non-mineralized rocks collected on traverses across mafic/ultramafic intrusions; the mineralized horizons occur at cross-overs in ratios such as Pd/Cu, Pd/Se and Pd/S.