During 2002-2004, the U.S. Geological Survey's National Water-Quality Assessment Program conducted a study to determine the effects of urbanization on stream water quality and aquatic communities in six environmentally heterogeneous areas of the conterminous United States--Atlanta, Georgia; Raleigh-Durham, North Carolina; Milwaukee-Green Bay, Wisconsin; Dallas-Fort Worth, Texas; Denver, Colorado; and Portland, Oregon. This report compares and contrasts the response of stream chemistry during base flow to urbanization in different environmental settings and examines the relation between the exceedance of water-quality benchmarks and the level of urbanization in these areas. Chemical characteristics studied included concentrations of nutrients, dissolved pesticides, suspended sediment, sulfate, and chloride in base flow. In three study areas where the background land cover in minimally urbanized basins was predominantly forested (Atlanta, Raleigh-Durham, and Portland), urban development was associated with increased concentrations of nitrogen and total herbicides in streams. In Portland, there was evidence of mixed agricultural and urban influences at sites with 20 to 50 percent urban land cover. In two study areas where agriculture was the predominant background land cover (Milwaukee-Green Bay and Dallas-Fort Worth), concentrations of nitrogen and herbicides were flat or decreasing as urbanization increased. In Denver, which had predominantly shrub/grass as background land cover, nitrogen concentrations were only weakly related to urbanization, and total herbicide concentrations did not show any clear pattern relative to land cover - perhaps because of extensive water management in the study area. In contrast, total insecticide concentrations increased with increasing urbanization in all six study areas, likely due to high use of insecticides in urban applications and, for some study areas, the proximity of urban land cover to the sampling sites. Phosphorus concentrations increased with urbanization only in Portland; in Atlanta and Raleigh-Durham, leachate from septic tanks may have increased phosphorus concentrations in basins with minimal urban development. Concentrations of suspended sediment were only weakly associated with urbanization, probably because this study analyzed only base-flow samples, and the bulk of sediment loads to streams is transported in storm runoff rather than base flow. Sulfate and chloride concentrations increased with increasing urbanization in four study areas (Atlanta, Raleigh-Durham, Milwaukee-Green Bay, and Portland), likely due to increasing contributions from urban sources of these constituents. The weak relation between sulfate and chloride concentrations and urbanization in Dallas-Fort Worth and Denver was likely due in part to high sulfate and chloride concentrations in ground-water inflow, which would have obscured any pattern of increasing concentration with urbanization. Pesticides often were detected at multiple sites within a study area, so that the pesticide 'signature' for a given study area - the mixtures of pesticides detected, and their relative concentrations, at streams within the study area - tended to show some pesticides as dominant. The type and concentrations of the dominant pesticides varied markedly among sites within a study area. There were differences between pesticide signatures during high and low base-flow conditions in five of the six study areas. Normalization of absolute pesticide concentrations by the pesticide toxicity index (a relative index indicating potential toxicity to aquatic organisms) dramatically changed the pesticide signatures, indicating that the pesticides with the greatest potential to adversely affect cladocerans or fish were not necessarily the pesticides detected at the highest concentrations. In a screening-level assessment, measured contaminant concentrations in individual base-flow water samples were compared with various water-qual