Resurgent cauldrons are defined as cauldrons (calderas) in which the cauldron block, following subsidence, has been uplifted, usually in the form of a structural dome. Seven of the best known resurgent cauldrons are: Valles, Toba, Creede, San Juan, Silverton, Lake City, and Timber Mountain. Geologic summaries of these and Long Valley, California, a probable resurgent caldera, are presented. Using the Valles caldera as a model, but augmented by information from other cauldrons, seven stages of volcanic, structural, sedimentary, and plutonic events are recognized in the development of resurgent cauldrons. They are: (I) Regional tumescence and generation of ring fractures; (II) Calderaforming eruptions; (III) Caldera collapse; (IV) Preresurgence volcanism and sedimentation; (V) Resurgent doming; (VI) Major ring-fracture volcanism; (VII) Terminal solfatara and hot-spring activity. These stages define the terminal cycle of resurgent cauldrons, which in the Valles caldera spanned more than 1 million years. The known and inferred occurrence of the seven stages in the eight cauldrons discussed, together with some time control in four cauldrons, indicates that resurgent doming is early in the postcollapse history; hence, it seems part of a pattern and not fortuitous. Doming of the cauldron block by magma pressure is preferred to doming by stock or laccolithic intrusion, although these processes may be subsidiary. Magma rise that produces doming may be explained in several ways, but the principal cause is not known. Nor is it known why some otherwise similar calderas do not have resurgent domes, although size and thickness of the cauldron block and the degree to which it was deformed during caldera collapse may be factors. All known resurgent structures are larger than 8 miles in diameter and are associated with silicic and, presumably, high-viscosity magmas. Genetically, resurgent cauldrons belong to a cauldron group in which subsidence of a central mass takes place along ring fractures and is related to eruption of voluminous ash flows, thereby differing from Kilauean-type calderas. It is proposed that typical Krakatoan-type calderas differ in that collapse is chaotic and ring fractures are not essential to their formation. Krakatoan calderas typically occur in the andesitic volcanoes of island arcs or the eugeosynclinal environment, and their sub-volcanic analogues are not known, whereas resurgent and related Glen Coe-type cauldrons are more common in cratonic or post-orogenic environments as are their sub-volcanic analogues - granitic ring complexes. Granitic ring complexes, such as Lirue, Sande, Ossipee, and Alnsj0, are probably the closest sub-volcanic analogues of resurgent calderas. The source areas of most of the ash-flow sheets of western United States and Mexico are yet to be found. It is suggested that many of them will prove to be resurgent structures. Present evidence suggests that ore deposits are more commonly associated with resurgent cauldrons than with other cauldron types.