The use of data analytics and digital mineral prospectivity mapping has become increasingly important to mineral resource assessment in recent years. Necessary information for many forms of such modeling are expert-validated training sets. Previous work defined occurrence and grade and tonnage models for Climax-type porphyry molybdenum deposits (Singer and others, 1986; Ludington and Plumlee, 2009), but these models contain less than 20 examples, which is not enough for a robust training set. The data files presented here define a broader training set for Climax-type porphyry molybdenum mineral systems, based primarily on the geochemical characteristics of the magmas that are the source of these deposits. Burt and others (1982) identified a special class of bimodal high-silica rhyolite lavas and shallow intrusions that they termed topaz rhyolites. These authors inferred the presence of high concentrations of fluorine in the magmas, identified the chemical similarity of the melts to those associated with Climax-type molybdenum porphyry deposits, and reported that the rocks were characterized by unusually high amounts of Cs, Rb, Li, U, Th, Nb, Ta, Sn, W, Mo, and Be. Burt and others (1982) identified 22 sites where these rhyolites occur, a list that would grow to 30 in subsequent publications (Christiansen and others, 1986; Christiansen and others, 2007). The connection between extreme petrologic evolution and the occurrence of Climax-type deposits was known at the time (Mutschler and others, 1978), but the availability of chemical data, particularly of trace-element data, made further progress at the time untenable. Today, the abundance of online data has made the next step in characterization of Climax-type magmas feasible. To create the database of 125 sites compiled in this data release, a number of online databases and other large collections of petrologic data were consulted, including NAVDAT (https://www.navdat.org/index.cfm), EARTHCHEM (https://www.earthchem.org/), GEOROC (https://georoc.mpch-mainz.gwdg.de/georoc/), and the National Geochemical Database of the U.S. Geological Survey (https://www.usgs.gov/centers/gggsc/science/national-geochemical-database). The initial filter used to select analyses of interest was Rb >300 ppm and Nb >30 ppm. During subsequent amplification and editing, these criteria were slightly relaxed in some cases. The result is a catalogue of sites where highly evolved rhyolites similar to those that occur in Climax-type mineral systems are known to occur. This catalogue should be a robust training set for the study of the distribution in time and space of Climax-type porphyry mineral systems. For each site, references were compiled that describe the geologic context, including the age and associated mineral deposits, for these highly evolved rhyolites.