A numerical ground-water flow model of a hypothetical basin was constructed and used to investigate the effects of ground-water withdrawals on rates of natural discharge to streams and springs in small basins of the Puget Sound Lowland. Definitions of the topography, geology, drainage, and climate of the hypothetical basin were based on the features of typical small basins in the Puget Sound Lowland. This information was used to construct a 13-layer numerical ground-water flow model capable of simulating water levels, hydraulic gradients, and discharge to streams and springs. Three sequences of glacial drift and interglacial deposits were simulated in the model; each sequence consisted of recessional outwash, till, advance, outwash, and fine-grained interglacial sediments. Alluvial sediments of the major stream valleys and undifferentiated glacial and interglacial deposits were also included in the model. The model was calibrated by comparing simulated hydrologic conditions with expected conditions and making adjustments to values of hydraulic characteristics as needed. The model was calibrated to predevelop- ment conditions (those prior to pumping), and then used to simulate the effects of pumping on natural discharge to streams and springs. Seven series of simulations were made to investigate the effects of (1) distance from the well to a stream, (2) the presence of confining layers, (3) pumping rate, (4) depth of the pumped aquifer, (5) distance from the well to a bluff, (6) well density, and (7) recharge rate. The discharge of wells pumping from unconfined outwash aquifers on the drift plains is derived almost entirely from capture of natural discharge to nearby stream reaches. Increasing the lateral distance between the well and stream caused more of the well discharge to be captured from other streams on the drift plain. Pumping from aquifers separated from the stream by one or more confining layers caused a reduction in the effects of pumping on discharge to nearby streams that was offset by an increase in the effects on discharge to more distant streams and springs. The percentage of well discharge captured from springs on the bluff was sensitive to the distance of wells from the bluff. Simulations also showed that increased well density caused greater water-level decline locally, but, at equilibrium, did not affect the extent of the area affected by reduction of natural discharge to streams and springs. Finally, decreased recharge in areas where development had created impervious surfaces had a direct effect on the natural discharge rates to streams and springs. Increased recharge, however, increased natural discharge and offset the effects of well withdrawals. Further analysis of the time-dependent effects of with- drawals would provide additional insights, but would require the development of a transient version of the model.