This report presents the results of a study of the geohydrology of stratified drift and streamflow in the Deerfield River Basin, northwestern Massachusetts. Detailed hydrologic information is needed to plan for the optimal use of ground-water and surface-water resources and for development of new drinking-water supplies in the basin. Sources and percentage of water available for recharge on an annual basis from October 1993 to September 1994, to the fine-grained stratified-drift in a narrow valley bordered by upland till and bedrock were: (1) direct infiltration of precipitation on the valley (30 percent); (2) tributary loss from an upland brook as it crosses the valley (7 percent); and (3) ground- and surface-water runoff from the uplands (63 percent). Seventy percent of recharge was available from upland sources. Seasonal variation in recharge caused changes in ground-water levels and flow directions. In early spring, the direction of flow is toward the valley axis, but in late summer, the direction of flow is nearly parallel to the valley axis. Field observations and results of a ground-water flow simulation indicated that water available for recharge was greater than actual recharge during the spring snowmelt and during intense precipitation events. In 1994, estimates of water available for recharge were greater than actual recharge by 10 percent in March and by 60 percent in April; actual recharge to the valley on an annual basis from October 1993 to September 1994 was 20 percent less than original estimates. A map showing thickness of stratified drift in the Connecticut Valley Lowlands indicates a deep north-south trending buried valley. Maximum thickness of the stratified drift is 385 feet. Interpretation of a seismic-reflection survey indicates fine-grained stratified drift may be underlain by coarse-grained deposits ranging in thickness from 0 to 150 feet. Hydraulic properties of the stratified drift were calculated from ground-water-level fluctuations induced by river stage changes using a ground-water-flow model for a site adjacent to the Deerfield River. A comparison of measured and simulated heads resulted in a vertical riverbed hydraulic conductivity of 3 feet per day, anisotropic ratio of horizontal to vertical hydraulic conductivity of 40:1, and storage of 0.040 and 0.0002 for the unconfined and confined layers of the stratified drift. Hydraulic diffusivity (transmissivity divided by the unconfined storage) at the site is about 168,000 feet squared per day. Streamflows at times of low flow were determined for 27 sites that drain areas ranging from 0.57 to 15.8 percent stratified drift. Streamflows exceeded between 80 and 99 percent of the time were determined for sites on the unregulated tributaries to the Deerfield River. Streamflows per square mile of drainage area were greatest from sites at the downstream ends of the North River-Colrain and the Deerfield River-Charlemont stratified-drift valleys. Flow-duration curves for three continuous streamflow-gaging stations on the regulated Deerfield River were compared to flow-duration curves for three continuous streamflow-gaging stations on unregulated tributaries to show the effects of dam regulation on streamflow. Flow- duration curves constructed using instantaneous discharges for the three regulated gaging stations have flat sections that correspond to the predominant streamflows when water is being released from storage from the dams.