<p>This study tests the efficacy of commonly used chemical fertility indicators in apatite and zircon for distinguishing Cu–Au porphyry intrusions of the Ordovician–Silurian Macquarie Arc from unmineralised I-type intrusions of the adjacent Silurian-Devonian Lachlan Orogen (south-eastern Australia). Chemical data were collected from apatite, zircon, and apatite inclusions in zircon, allowing the integration of petrological information from these two common accessory minerals. Melt Cl and F estimates, determined with available mineral-melt partitioning models for apatite, indicate that the precursor melts to fertile intrusions of the Macquarie Arc were generally not elevated in Cl relative to unmineralised intrusions, or to arc-related melts more generally. Melt oxidation sensitive trace elements in zircon (Ce) and apatite (S, Mn and Fe) suggest that porphyry Cu–Au fertile melts of the Macquarie Arc were not invariably more oxidised than unmineralised I-type granitic intrusions. Apatite Sr and Y concentrations and Eu/Eu* and Dy/Yb in zircon offer evidence of hornblende stability and delayed plagioclase crystallisation in most Cu–Au fertile melts of the Macquarie Arc, supporting a critical role for elevated water content in determining the metallogenic potential of arc-related magmas. Evidence of dynamic melt hydration and oxidation conditions are preserved in zircon and apatite populations of the fertile intrusions. We find that devolatilisation may affect ore fertility indicators in apatite and zircon to the extent that these signatures become indistinguishable from unmineralised intrusions of a similar composition. Zircon fertility indictors are found to be less effective at discriminating alkalic (shoshonitic) porphyries associated with Au mineralisation from infertile arc igneous suites, but these intrusions can instead be distinguished by high Sr/Y in apatite.</p>