Deep geothermal drilling in the resurgent dome of the Valles caldera has allowed us to define a consistent intracaldera stratigraphic sequence that differs in a number of respects from the temporally equivalent sequence outside the caldera. Above the deeply eroded, Pliocene Paliza Canyon Formation, felsic ash flows and sediments in the dome area form a complex sequence of undetermined age that we call the Lower Tuffs. An erosional interval separates these rocks from the overlying Otowi Member (1.4 m.y.) of the Bandelier Tuff. Another period of erosion, during which a tuffaceous sandstone was deposited, separates the Ottowi from the overlying Tshirege Member (1.1 m.y.) of the Baldelier. Both the Otowi and Tshirege members, with maximum thicknesses of 833 and 1155 m, respectively, are substantially thicker within the caldera than outside, indicating simultaneous deposition and cauldron subsidence. Both are predominantly densely welded with distinction interior zones and granophyric crystallization. Resurgent doming was initiated after emplacement of the Tshirege Member as evidenced by erosion of the upper portions of the cooling unit prior to deposition of overlying units. An isopach map of the sandstone deposited during this erosional period shows that streams draining the uplifting dome were localized along the present Redondo Creek trend. Subsequent volcanic activity resulted in the formation of at least three additional ash flow tuff cooling units prior to deposition of caldera fill and eruption of the Redondo Creek Member of the Valles Rhyolite. A numerical model applied to formation of the Redondo dome suggests that the top of the Causative magma body is located at a depth of about 4700 m, beneath the bottom of the deepest geothermal well in the dome. No wells have penetrated intrusives that could be related to this magma. We suggest that the locations of faults bounding the apical graben of the dome were influenced by older faults associated with the northeast trending Jemez lineament. These faults were active early in the uplift history of the dome and account for many of the structural differences between hypotheical dome development and reality.