Rhyolites of the Picabo volcanic field (10.4–6.6 Ma) in eastern Idaho are preserved as thick ignimbrites and lavas along the margins of the Snake River Plain (SRP), and within a deep (>3 km) borehole near the central axis of the Yellowstone hotspot track. In this study we present new O and Hf isotope data and U–Pb geochronology for individual zircons, O isotope data for major phenocrysts (quartz, plagioclase, and pyroxene), whole rock Sr and Nd isotope ratios, and whole rock geochemistry for a suite of Picabo rhyolites. We synthesize our new datasets with published Ar–Ar geochronology to establish the eruptive framework of the Picabo volcanic field, and interpret its petrogenetic history in the context of other well-studied caldera complexes in the SRP. Caldera complex evolution at Picabo began with eruption of the 10.44±0.27 Ma (U–Pb) Tuff of Arbon Valley (TAV), a chemically zoned and normal-δ18O (δ18O magma=7.9‰) unit with high, zoned 87Sr/86Sri (0.71488–0.72520), and low-εNd(0) (−18) and εHf(0) (−28). The TAV and an associated post caldera lava flow possess the lowest εNd(0) (−23), indicating ∼40–60% derivation from the Archean upper crust. Normal-δ18O rhyolites were followed by a series of lower-δ18O eruptions with more typical (lower crustal) Sr–Nd–Hf isotope ratios and whole rock chemistry. The voluminous 8.25±0.26 Ma West Pocatello rhyolite has the lowest δ18O value (δ18Omelt=3.3‰), and we correlate it to a 1,000 m thick intracaldera tuff present in the INEL-1 borehole (with published zircon ages 8.04–8.35 Ma, and similarly low-δ18O zircon values). The significant (4–5‰) decrease in magmatic-δ18O values in Picabo rhyolites is accompanied by an increase in zircon δ18O heterogeneity from ∼1‰ variation in the TAV to >5‰ variation in the late-stage low-δ18O rhyolites, a trend similar to what is characteristic of Heise and Yellowstone, and which indicates remelting of variably hydrothermally altered tuffs followed by rapid batch assembly prior to eruption. However, due to the greater abundance of low-δ18O rhyolites at Picabo, the eruptive framework may reflect an intertwined history of caldera collapse and coeval Basin and Range rifting and hydrothermal alteration. We speculate that the source rocks with pre-existing low-δ18O alteration may be related to: (1) deeply buried and unexposed older deposits of Picabo-age or Twin Falls-age low-δ18O volcanics; and/or (2) regionally-abundant late Eocene Challis volcanics, which were hydrothermally altered near the surface prior to or during peak Picabo magmatism. Basin and Range extension, specifically the formation of metamorphic core complexes exposed in the region, could have facilitated the generation of low-δ18O magmas by exhuming heated rocks and creating the large water-rock ratios necessary for shallow hydrothermal alteration of tectonically (rift zones) and volcanically (calderas) buried volcanic rocks. These interpretations highlight the major processes by which supereruptive volumes of magma are generated in the SRP, mechanisms applicable to producing rhyolites worldwide that are facilitated by plume driven volcanism and extensional tectonics.