High-silica, topaz-bearing rhyolites of Cenozoic age are widely distributed across the western United States and Mexico. Topaz rhyolites are characteristically enriched in fluorine (>0.2 wt%) and contain topaz crystallize during post-magmatic vapor-phase alteration. In the United States, their ages span much of the Cenozoic Era (50 to 0.06 Ma). Their emplacement followed or was contemporaneous with calc-alkaline and basaltic magmatism in the Basin and Range province, along the Rio Grande rift, and in Montana, and coincided with episodes of extensional tectonism in these regions. Nearly all topaz rhyolites extruded as small, endogenous lava domes with or without lava flows; no topaz-bearing ash-flow tuffs have yet been identified with certainty in the western United States. Most domes are underlain by a precursory blanket of non-welded tephra. A few are small, shallowly emplaced intrusive plugs. Volumes of rock (3) in individual complexes composed of 1 to many separate extrusions suggest that the lavas were erupted from small to medium sized magma bodies. In addition to topaz, these rhyolites also contain garnet, bixbyite, pseudobrookite, hematite, and fluorite in cavities or in their devitrified groundmasses. All of these phases may form during vapor-phase crystallization. Magmatic phenocrysts include sanidine (ca. Or50), quartz, sodic plagioclase (usually oligoclase), and F- and Fe-rich biotite in order of usual abundance. Fe-rich hornblende or clinopyroxene occur in a few lavas. Common magmatic accessory minerals include magnetite, ilmenite, zircon, apatite, allanite, and fluorite. Titanite and REE-rich phosphates have been identified in a few lavas. The rhyolites crystallized over a wide temperature interval (850 to 600°C, with most at the lower end of this range) and at variable oxygen fugacities. Titanite-bearing lavas crystallized above the NNO buffer under oxidizing conditions. Most others appear to have crystallized near the QFM oxygen buffer. For individual complexes, temperatures correlate negatively with F-content. All topaz rhyolites are high-SiO2 rhyolites with elevated F, Na, K, Fe/Mg and low Ti, Mg, Ca, and P. Samples with F concentrations of about 1% have notably lower Si and higher Al and Na than other topaz rhyolite glasses. Most glasses from topaz rhyolites are metaluminous, but many appear to be slightly peraluminous. Fluorine concentrations in glasses range from slightly less than 0.2 to more than 1.0 wt%, and F/Cl ratios are high (3 to 10) compared to F-rich peralkaline glasses (N = 1 to 3) and have pronounced negative Eu anomalies (Eu/Eu* = 0.01 to 0.02). Both of these parameters decrease with differentiation as indicated by increasing F, U, Cs, and other incompatible elements. Titanite-bearing rhyolites have prominent middle REE depletions. Initial Sr-isotope ratios range from 0.705 to over 0.710. Geochemical trends at individual complexes are interpreted as arising from fractional crystallization of an initially more “mafic” rhyolite with about 0.2% fluorine. Extensive fractionation of sanidine, quartz, plagioclase, biotite, and Fe-Ti oxides (in proportions consistent with their modes) produced much of the trace element patterns. Zircon, apatite, and a REE-rich phase (allanite, monazite, or titanite) were minor but important fractionating phases. No liquid-state fractionation is required to explain the geochemical trends. The high F content and F/Cl ratios of topaz rhyolite melts may have modified phase relations so as to produce Na and Al enrichments for highly evolved magmas. Topaz rhyolites are intimately related to economic deposits of lithophile elements (i.e. Be, U, F, Li, and Sn). The volcanic rocks were probably ore- and, in some cases, fluid-sources for these mineral deposits. In their age, tectonic setting, mineralogy, chemistry, and style of emplacement, topaz rhyolites bear resemblance to the rhyolitic stocks associated with Climax-type Mo deposits, and some may be surface manifestations of such deposits. In their chemical composition and mineralogy, topaz rhyolites are distinct from both peralkaline rhyolites and calc-alkaline rhyolites with which they may be spatially and temporally associated. Some of the compositional differences between topaz rhyolites and peralkaline rhyolites may be attributed to the relative effects of F and Cl, on melt structure and phase relationships in their parental magmas. The F/Cl ratios of the melt or its source may determine the alumina saturation of the magma series. Topaz rhyolites are distinguishable from calc-alkaline rhyolites by lower Sr, Ba, and Eu, and higher F, Rb, U, and Th. The usually low La/Yb ratios of topaz rhyolites distinguish them from both peralkaline and calc-alkaline rhyolite suites. Topaz rhyolites are similar to other aluminous rhyolites erupted in bimodal associations with basalt in the western United States. They may be the equivalent of the topaz-bearing ongonites of central Asia. Topaz rhyolites from the western United States are not the eruptive equivalents of S-type granites; we liken them to the highly evolved, non-peralkaline, and F-rich anorogenic granites. Topaz rhyolites appear to have evolved from partial melts of a residual felsic granulite source in the lower or middle crust of the Precambrian continent. According to the proposed model, the passage of contemporaneous mafic magmas through the crust produced necessarily small volumes of partial melts as a result of the decomposition of small amounts of F-rich biotite that persisted in a high-grade metamorphic protolith. An extensional tectonic setting allowed these small batches of magma to rise without substantial mixing with contemporaneous mafic magmas. Subsequent fractionation led to their extreme trace element characteristics.