Abstract Magnetite-bearing cavities in rhyolitic effusives and sulfide-bearing miarolithic cavities in acid intrusives are taken as natural models of epigenetic Fe—Cu—Zn—Pb ores. Corresponding mineral equilibria and some of their limitations are reviewed. Mineral sequences suggest that pH varied during mineral formation as from 12–9 above 300° C to about 6 at 100° C. Paragenetical considerations indicate that halogen acids are not essential in the formation of Fe—Cu—Zn—Pb ores. The role of carbonic acid is discussed and a hypothesis on the critical role of carbonate-complexes is advanced. It is pointed out that any formation process involving decomplexing and precipitation would give rise to replacement relations or the type seen in ore minerals, and that prior to mineral deposition, ore-solutions are unsaturated in ore constituents and therefore able to extract them from traversed wallrock.