A numerical-modeling study was done to better understand hydrologic-system responses to ground-water withdrawals in the Hunt-Annaquatucket-Pettaquamscutt (HAP) stream-aquifer system of Rhode Island. System responses were determined by use of steady-state and transient numerical ground-water-flow models. These models were initially developed in the late 1990s as part of a larger study of the stream-aquifer system. The models were modified to incorporate new data made available since the original study and to meet the objectives of this study. Changes made to the models did not result in substantial changes to simulated ground-water levels, hydrologic budgets, or streamflows compared to those calculated by the original steady-state and transient models. Responses of the hydrologic system are described primarily by changes in simulated streamflows and ground-water levels throughout the basin and by changes to flow conditions in the aquifer in three wetland areas immediately east of the Lafayette State Fish Hatchery, which lies within the Annaquatucket River Basin in the town of North Kingstown. Ground water is withdrawn from the HAP aquifer at 14 large-capacity production wells, at an industrial well, and at 3 wells operated by the Rhode Island Department of Environmental Management at the fish hatchery. A fourth well has been proposed for the hatchery and an additional production well is under development by the town of North Kingstown. The primary streams of interest in the study area are the Hunt, Annaquatucket, and Pettaquamscutt Rivers and Queens Fort Brook. Total model-calculated streamflow depletions in these rivers and brook resulting from withdrawals at the production, industrial, and fish-hatchery wells pumping at average annual 2003 rates are about 4.8 cubic feet per second (ft3/s) for the Hunt River, 3.3 ft3/s for the Annaquatucket River, 0.5 ft3/s for the Pettaquamscutt River, and 0.5 ft3/s for Queens Fort Brook. The actual amount of streamflow reduction in the Annaquatucket River caused by pumping actually is less, 1.1 ft3/s, because ground water that is pumped at the fish-hatchery wells (2.2 ft3/s) is returned to the Annaquatucket River after use at the hatchery. One of the primary goals of the study was to evaluate the response of the hydrologic system to simulated withdrawals at the proposed well at the fish hatchery. Withdrawal rates at the proposed well would range from zero during April through September of each year to a maximum of 260 gallons per minute [about 0.4 million gallons per day (Mgal/d)] in March of each year. The average annual withdrawal rate at the fish hatchery resulting from the addition of the proposed well would increase by only 0.13 ft3/s, or about 5 percent of the 2003 withdrawal rate. The increased pumping rate at the hatchery would further reduce the average annual flow in Queens Fort Brook by less than 0.05 ft3/s and in the Annaquatucket River by about 0.1 ft3/s (which includes some model error). A new production well in the Annaquatucket River Basin is under development by the town of North Kingstown. A simulated pumping rate of 1.0 Mgal/d (1.6 ft3/s) at this new well resulted in additional streamflow depletions, compared to those calculated for the 2003 withdrawal conditions, of 0.8 and 0.2 ft3/s in the Annaquatucket and Pettaquamscutt Rivers, respectively. The source of water for about 30 percent of the well's pumping rate, or about 0.5 ft3/s, is derived from ground-water inflow from the Chipuxet River Basin across a natural ground-water drainage divide that separates the Annaquatucket and Chipuxet River Basins; the remaining 0.1 ft3/s of simulated pumping consists of reduced evapotranspiration from the water table. Model-calculated changes in water levels in the aquifer for the various withdrawal conditions simulated in this study indicate that ground-water-level declines caused by pumping are generally less than 5 feet (ft). However, ground-water-level declines of as