In 2005, about 12 percent of total water withdrawals (440 million gallons per day (Mgal/d)) in Vermont were from groundwater sources (51 Mgal/d), and about 88 percent were from surface-water sources (389 Mgal/d). Of total water withdrawals, about 78 percent were used for cooling at a power plant, 9 percent were withdrawn by public suppliers, about 5 percent were withdrawn for domestic use, about 3 percent were withdrawn for use at fish hatcheries, and the remaining 5 percent were divided among commercial/industrial, irrigation, livestock, and snowmaking uses. About 49 percent of the population of Vermont was supplied with drinking water by a public supplier, and 51 percent was self supplied. Some of the Minor Civil Divisions (MCDs) that had large self-supplied populations were located near the major cities of St. Albans, Burlington, Montpelier, Barre, and Rutland, where the cities themselves were served largely by public supply, but the surrounding areas were not. Most MCDs where withdrawals by community water systems totaled more than 1 Mgal/d used predominantly surface water, and those where withdrawals by community water systems totaled 1 Mgal/d or less used predominantly groundwater. Withdrawals of groundwater greater than 1 Mgal/d were made in Middlebury, Bethel, Hartford, Springfield, and Bennington, and withdrawals of surface water greater than 2 Mgal/d were made in Grand Isle, Burlington, South Burlington, Mendon, Brattleboro, and Vernon. Increases in groundwater withdrawals greater than 0.1 Mgal/d are projected for 2020 for Fairfax, Hardwick, Middlebury, Sharon, Proctor, Springfield, and Manchester. The largest projected increases in surface-water withdrawals from 2005 to 2020 are located along the center axis of the Green Mountains in the ski-area towns of Stowe, Warren, Mendon, Killington, and Wilmington. In 2005, withdrawals were at least 1 Mgal/d greater than return flows in South Burlington, Waterford, Orange, Mendon, Woodford, and Vernon. Many of these MCDs had small populations themselves but provided water to community water systems in neighboring towns or cities. Wilmington probably will be added to this list by 2020 because of proposed new withdrawals for snowmaking in Dover. About 15 percent of MCDs had greater return flows than withdrawals; possible reasons are water importation, larger service areas for municipal sewer than for municipal water resulting in underestimation of withdrawals, leakage into sewer pipes, faulty assumptions in assigning coefficients, or other limitations of the study methods. To store and facilitate retrieval of water-use estimates and data for 2005 and projections for 2020, a water-use database for Vermont was designed and populated. Data include withdrawals and return flows from and to groundwater and surface water for all individual facilities and entities that are in Vermont drinking water, discharge permit, or other State water-use databases, along with estimates for many other facilities. Also included are estimates for aggregated domestic and livestock withdrawals and return flows by census block. Retrievals from the database and summaries presented in this report can be used to help identify areas where projected growth in Vermont from 2005 to 2020 might affect groundwater availability.