The MMH porphyry type copper–molybdenum deposit in northern Chile is the newest mine in the Chuquicamata District, one of largest copper concentrations on Earth. Mineralized Eocene–Oligocene porphyry intrusions are hosted by essentially barren Triassic granodiorites. Despite a century of exploitation, geologists still have problems in the mine distinguishing the Triassic granodiorite from the most important ore-carrying Eocene porphyries in the district. To resolve the problem, internally consistent high-quality geochemical analyses of the Triassic and Tertiary intrusives were carried out: explaining the confusion, they show that the rock units in question are nearly identical in composition and thus respond equally to hydrothermal alteration. In detail, the only difference in terms of chemical composition is that the main Eocene–Oligocene porphyries carry relatively less Fe and Ni. Unexpectedly, the mineralized Eocene–Oligocene porphyries have consistently less U and Th than other Tertiary intrusions in the district, a characteristic that may be valuable in exploration. The supergiant copper–molybdenum deposits in the Central Andes were formed within a narrow interval between 45 and 31 Ma, close to 7% of the 200 My duration of “Andean” magmatism, which resulted from subduction of oceanic lithosphere under South America since the Jurassic. Although recent work has shown that subduction was active on the margin since Paleozoic times, pre-Andean (pre-Jurassic) “Gondwanan” magmatism is often described as being very different, having involved crustal melting and the generation of massive peraluminous rhyolites and granites. This study shows that the indistinguishable Late Triassic and Eocene–Oligocene intrusions occupy the same narrow NS geographic belt in northern Chile. If it is accepted that magma character may determine the potential to generate economic Cu–Mo deposits, then Late Triassic volcano-plutonic centres in the same location in the South American margin could have contained valuable ore deposits, although their preservation will depend on the level attained by pre-mid Jurassic erosion. Both Late Triassic and Eocene–Oligocene magmatic events occurred during the waning stages of vigorous volcano-plutonic cycles, and both preceded apparent gaps in igneous activity (Rhaetian and post-Oligocene), abrupt lateral shifts of the volcanic front and radical changes in the character of the magmas generated. Both Late Triassic and Eocene–Oligocene intrusions were emplaced along the same narrow strip of crust; it is probable that they both exploited the same deep crustal structures. The Eocene–Oligocene magmatic front was controlled by an orogen-parallel shear system caused by oblique subduction; it is possible that Late Triassic magmatism along the same belt had a similar setting. The identified Rhaetian gap in subduction and magmatism may have widespread implications.