PGE mineralization has been identified in various rock types and at various stratigraphic levels in layered intrusions of any age, size and magmatic lineage, but the most important deposits occur as relatively narrow stratiform reefs in the lower to central ultramafic-mafic portions of large tholeiitic intrusions of late Archean to early Proterozoic age. One of the main challenges in exploration is that the reefs tend to be sulfide-poor. In many chromitites, magnetitites and silicate-hosted ores, the rocks contain no visible sulfides, possibly due to (late) magmatic sulfide resorption. As a result, some deposits may have been overlooked, particularly those in the upper portions of the intrusions that were in the past considered to be relatively unprospective. Amongst lithogeochemical tools, Cu/Pd ratios have proven to be particularly useful to evaluate the PGE potential of intrusions and to delineate the position of the reefs within the intrusions.The origin of the PGE mineralization remains controversial. A possible explanation for the low sulfide contents of many PGE-rich intrusions is that most of their parental magmas were strongly undersaturated in sulfur and at least partially derived from the Spoor and PGE-enriched sub-continental lithospheric mantle. Sulfide saturation upon emplacement in the crust may have been reached during differentiation. Empirical evidence supports theoretical considerations that chromite and magnetite precipitation may be particularly conducive to trigger sulfide melt saturation, due to a pronounced decrease in FeO content of the magma. The importance of magma mixing in triggering sulfide supersaturation remains unclear. The same applies to contamination; some intrusions show a distinct crustal component, but many others do not, and there is little if any correlation between sulfide content and crustal component. Together with the general paucity of sulfides in the intrusions this could suggest that contamination is not critical in reef formation and may indeed be a negative factor.Other processes may also be relevant to reef formation. Data from the well-studied Bushveld Complex suggest that the magmas had reached sulfide saturation prior to emplacement, and that sulfides were entrained in the magma during ascent and emplacement. Sulfide entrainment has previously been recognised as one of the key factors in the formation of massive Ni-Cu sulfide deposits, and it is suggested here that it is also relevant to the formation of PGE deposits. (c) 2005 Elsevier Ltd. All rights reserved.