The surface-water hydrology in Clear, Clarks, and Clover Creek Basins in central Pierce County, Washington, is described with a conceptual model of the runoff processes and then simulated with the Hydrological Simulation Program-FORTRAN (HSPF), a continuous, deterministic hydrologic model. The study area is currently undergoing a rapid conversion of rural, undeveloped land to urban and suburban land that often changes the flow characteristics of the streams that drain these lands. The complex interactions of land cover, climate, soils, topography, channel characteristics, and ground- water flow patterns determine the surface-water hydrology of the study area and require a complex numerical model to assess the impact of urbanization on streamflows. The U.S. Geological Survey completed this investigation in cooperation with the Storm Drainage and Surface Water Management Utility within the Pierce County Department of Public Works to describe the important rainfall-runoff processes within the study area and to develop a simulation model to be used as a tool to predict changes in runoff characteristics resulting from changes in land use. The conceptual model, a qualitative representation of the study basins, links the physical characteristics to the runoff process of the study basins. The model incorporates 11 generalizations identified by the investigation, eight of which describe runoff from hillslopes, and three that account for the effects of channel characteristics and ground-water flow patterns on runoff. Stream discharge was measured at 28 sites and precipitation was measured at six sites for 3 years in two overlapping phases during the period of October 1989 through September 1992 to calibrate and validate the simulation model. Comparison of rainfall data from October 1989 through September 1992 shows the data-collection period beginning with 2 wet water years followed by the relatively dry 1992 water year. Runoff was simulated with two basin models-the Clover Creek Basin model and the Clear-Clarks Basin model-by incorporating the generalizations of the conceptual model into the construction of two HSPF numerical models. Initially, the process-related parameters for runoff from glacial-till hillslopes were calibrated with numerical models for three catchment sites and one headwater basin where streamflows were continuously measured and little or no influence from ground water, channel storage, or channel losses affected runoff. At one of the catchments soil moisture was monitored and compared with simulated soil moisture. The values for these parameters were used in the basin models. Basin models were calibrated to the first year of observed streamflow data by adjusting other parameters in the numerical model that simulated channel losses, simulated channel storage in a few of the reaches in the headwaters and in the floodplain of the main stem of Clover Creek, and simulated volume and outflow of the ground-water reservoir representing the regional ground-water aquifers. The models were run for a second year without any adjustments, and simulated results were compared with observed results as a measure of validation of the models. The investigation showed the importance of defining the ground-water flow boundaries and demonstrated a simple method of simulating the influence of the regional ground-water aquifer on streamflows. In the Clover Creek Basin model, ground-water flow boundaries were used to define subbasins containing mostly glacial outwash soils and not containing any surface drainage channels. In the Clear-Clarks Basin model, ground-water flow boundaries outlined a recharge area outside the surface-water boundaries of the basin that was incorporated into the model in order to provide sufficient water to balance simulated ground-water outflows to the creeks. A simulated ground-water reservoir used to represent regional ground-water flow processes successfully provided the proper water balance of inflows and outfl