The modern Sierra Nevada and Great Basin were likely the site of a high-elevation orogenic plateau well into Cenozoic time, supported by crust thickened during Mesozoic shortening. Although crustal thickening at this scale can lead to extension, the relationship between Mesozoic shortening and subsequent formation of the Basin and Range is difficult to unravel because it is unclear which of the many documented or interpreted extensional episodes was the most significant for net widening and crustal thinning. To address this problem, we integrate geologic and geochronologic data that bear on the timing and magnitude of Cenozoic extension along an ???200km east-west transect south of Winnemucca, Battle Mountain, and Elko, Nevada. Pre-Cenozoic rocks in this region record east-west Palaeozoic and Mesozoic compression that continued into the Cretaceous. Little to no tectonism and no deposition followed until intense magmatism began in the Eocene. Eocene and Oligocene ash-flow tuffs flowed as much as 200km down palaeovalleys cut as deeply as 1.5km into underlying Palaeozoic and Mesozoic rocks in a low-relief landscape. Eocene sedimentation was otherwise limited to shallow lacustrine basins in the Elko area; extensive, thick clastic deposits are absent. Minor surface extension related to magmatism locally accompanied intense Eocene magmatism, but external drainage and little or no surface deformation apparently persisted regionally until about 16-17Ma. Major upper crustal extension began across the region ca. 16-17Ma, as determined by cross-cutting relationships, low-temperature thermochronology, and widespread deposition of clastic basin fill. Middle Miocene extension was partitioned into high-strain (50-100%) domains separated by largely unextended crustal blocks, and ended by 10-12Ma. Bimodal volcanic rocks that erupted during middle Miocene extension are present across most of the study area, but are volumetrically minor outside the northern Nevada rift. The modern physiographic basins and ranges formed during a distinctly different episode of extension that began after about 10Ma and has continued to the present. Late Miocene and younger faulting is characterized by widely spaced, high-angle normal faults that cut both older extended and unextended domains. Major widening of the Basin and Range at this latitude thus took place during a relatively brief interval in the middle Miocene, and the lack of major shortening west of the Sierra Nevada at this time suggests that the change in the plate margin from microplate subduction to lengthy transtensional strike-slip played an important role in allowing extension to occur when it did, as rapidly as it did. The onset of extension ca. 16-17Ma was coeval with both Columbia River flood-basalt volcanism and the hypothesized final delamination of the shallow Farallon slab that lay beneath the western USA in the early Tertiary. However, it is unclear if these events were necessary prerequisites for extension, simply coincidental, or themselves consequences of rapid extension and/or reorganization of the plate boundary.