In ancient volcano-sedimentary successions, the reconstruction of the facies architecture may be difficult because of incomplete exposure, deformation, metamorphism and hydrothermal alteration. The distinction of lavas or domes from shallow intrusions is critical for the ordering of volcanic, intrusive, sedimentary and hydrothermal events. Both lavas and shallow intrusions typically have porphyritic textures and may have breccia along their top contacts. Detailed study of these breccias and their contact relationships is a reliable means of distinguishing lavas or domes from shallow intrusions. In this contribution, we describe rhyolites in an Early Paleozoic, submarine, volcano-sedimentary succession in northwestern Argentina, known as the Chiquero Formation, formed in a rapidly subsiding backarc setting. This succession is mainly composed of turbidite interbedded with porphyritic rhyolite-dacite units.Many of the rhyolites have breccias along their upper contacts. Rhyolite clasts in the breccias match the underlying rhyolites in terms of phenocryst populations and texture. The main breccia facies along upper contacts are monomictic rhyolite breccia, mudstone-matrix rhyolite breccia and siliceous mudstone-matrix rhyolite breccia. The monomictic rhyolite breccia is interpreted to be in situ hyaloclastite formed by quench fragmentation of the tops of rhyolite lavas extruded onto the sea floor. The mudstone-matrix rhyolite breccia and the siliceous mudstone-matrix rhyolite breccia are also interpreted as hyaloclastite but the spaces between the rhyolite clasts have been infilled by the locally available sediment and hydrothermal silica, respectively. The mudstone-matrix rhyolite breccia and the siliceous mudstone-matrix rhyolite breccia superficially resemble peperite in being composed of igneous and sedimentary components. However, the infilling sediment is laminated and the laminae are conformable with bedding in the overlying turbidite sequence, implying that the sediment passively infilled pore spaces in pre-existing rhyolite breccia. Hence, the studied rhyolites of the Chiquero Formation were submarine lavas emplaced on the sea floor. Those lavas were probably emplaced during the Middle Ordovician based on correlation with a dated rhyolite interpreted to be syn-sedimentary. However, at some localities, the turbidite sequence contains Tremadocian/Arenigian (Early Ordovician) fossils and a discrepancy exists between the isotopic age of the rhyolites and the age of the host sedimentary sequence based on fossils. It is likely that the Chiquero Formation spans a wider age range than previously thought, from Early Ordovician to Late Ordovician, and that at least four episodes of rhyolitic-dacitic magmatism contributed to this backarc succession.