Wyoming has historically experienced extended periods of drought, which have had significant economic and social impacts. Tree-ring records and archeological evidence indicate that past centuries have contained multi-decadal "megadroughts" far more severe than any drought of the past 150 years. Here, we propose to study the extent and controls on past dry periods in the Bighorn and Green River watersheds that exceeded even the severity of multi-decadal "megadroughts." In doing so, we will build upon results from our previous Wyoming Water Research Program grant, which enabled us to measure the prehistoric shoreline elevations of a lake, Lake of the Woods, at the point where the watersheds of the Green and Bighorn Rivers meet with the watershed of the Snake River. Our results showed that climatic shifts can rapidly generate new hydrologic regimes that persist for centuries to millennia. Aridity at least as severe and extensive as during the CE 1930s Dust Bowl prevailed from 9.1-3.4 ka (thousand years before CE 1950), and was most extensive across the watersheds from ca. 5-3 ka. Regional aridity was also bounded by abrupt shifts at >11.3 and 1.8-1.2 ka. Other results from our previous grant indicated that the North Platter River probably had ephemeral flows for several millennia. These centennial-to-millennial variations in the availability of water are important because they occurred when the world's climate was similar to today, and represent part of the natural climate variability that preceded historic time. Such persistent drought may well represent an analogy for periods in Wyoming's future, and should be studied. Proposed work would include 1) an extensive survey of lakes in the Green and Bighorn drainage basins, using sub-surface radar, to determine the extensiveness of past periods of low lake levels, 2) sediment core analysis, including radiocarbon dating and fossil pollen analyses, of a few representative lakes to date and quantify past climate conditions, and 3) hydroclimatic analysis, comparing paleoclimate estimates with modern climatic data and climate model output, to examine the factors that contributed to the periods of prolonged drought. Improved data would provide direct measurements of regional water levels during extremely dry periods and, thus, could provide important data for considering future drought impacts and the development of contingency plans.