Areas contributing recharge and sources of water to one proposed and seven present public-supply wells, screened in sand and gravel deposits and clustered in three study areas, were determined on the basis of calibrated, steady-state ground-water-flow models representing average hydrologic conditions. The area contributing recharge to a well is defined as the surface area where water recharges the ground water and then flows toward and discharges to the well. In Cumberland and Lincoln, public-supply well fields on opposite sides of the Blackstone River are in a narrow valley bordered by steep hillslopes. Ground-water-level and river-stage measurements indicated that river water was infiltrating the aquifer and flowing toward the wells during pumping conditions. Simulated areas contributing recharge to the Cumberland well field operating alone for both average (324 gallons per minute) and maximum (1,000 gallons per minute) pumping rates extend on both sides of the river to the lateral model boundaries, which is the contact between the valley and uplands. The area contributing recharge at the average pumping rate is about 0.05 square mile and the well field derives 72 percent of pumped water from upland runoff. At the maximum pumping rate, the area contributing recharge extends farther up and down the valley to 0.12 square mile and the primary source of water to the well field was infiltrated river water (53 percent). Upland areas draining toward the areas contributing recharge encompass 0.58 and 0.66 square mile for the average and maximum rates, respectively. By incorporating the backup Lincoln well-field withdrawals (2,083 gallons per minute) into the model, the area contributing recharge to the Cumberland well field operating at its maximum rate is reduced to 0.08 square mile; part of the simulated area which contributes recharge to the Cumberland well field when it is operating alone contributes instead to the Lincoln well field when both well fields are pumped. The Cumberland well field compensates by increasing the percentage of water it withdraws from the river by 11 percent. The upland area draining toward the Cumberland contributing area is 0.55 square mile. The area contributing recharge to the Lincoln well field is 0.08 square mile and infiltrated river water contributes 88 percent of the total water; the upland area draining toward the contributing area is 0.34 square mile. In North Smithfield, a public-supply well in a valley-fill setting is close to Trout Brook Pond, which is an extension of the Lower Slatersville Reservoir. A comparison of water levels from the pond and underlying sediments indicates that water is not infiltrated from Trout Brook Pond when the supply well is pumped at its maximum rate of 200 gallons per minute. Simulated areas contributing recharge for the maximum pumping rate and for the estimated maximum yield, 500 gallons per minute, of a proposed replacement well extend to the ground-water divides on both sides of Trout Brook Pond. For the 200 gallons-per-minute rate, the area contributing recharge is 0.23 square mile; the well derives almost all of its water from intercepted ground water that normally discharges to surface-water bodies. For the pumping rate of 500 gallons per minute, the area contributing recharge is 0.45 square mile. The increased pumping rate is balanced by additional intercepted ground water and by inducing 25 percent of the total withdrawn water from surface water. In Westerly, one public-supply well is in a watershed where the primarily hydrologic feature is a wetland. Water levels in piezometers surrounding the well site indicated a downward vertical gradient and the potential for water in the wetland to infiltrate the underlying aquifer. The simulated area contributing recharge for the average pumping rate (240 gallons per minute) and for the maximum pumping rate (700 gallons per minute) extends to the surrounding uplands (surficial materials not covered by t