We are building on our experience integrating soil surveys and soil laboratory data for sampled sites in the Western Oregon Cascades and the Oregon Coast Range to estimate the spatial variation of soil C, N, water retention, and hydraulic conductivity for larger portions of the Columbia River Basin. This information is being utilized in the ecosystem process model GEM-TOPMODEL, to estimate the movement of nutrients and pollutants through soils, which influence both terrestrial and aquatic biota. Our approach uses detailed digital soil mapping (SSURGO) and state-level mapping (STATSGO) to spatially extrapolate soil properties derived from the National Soil Characterization Database (NSCD) using soil classification as the link. Many SSURGO and STATSGO soil properties are given a range of values and the mean is not necessarily representative of the actual properties, so including data from the NSCD provides a means of quantifying soil property variation. Soil classification for the NSCD, SSURGO, and STATSGO was frequently modified to make it consistent with the current version of Soil Taxonomy since it plays a central role in linking point and mapping data. The current classification for named series was assigned using the national soil series classification (SC) database and obsolete soil taxa lacking series names were changed to current taxa using relationships based on changes from the 1982 Natural Resources Inventory compared with the SC database from 2005. Bulk density (BD) was often missing from the NSCD, and because BD is crucial for making spatial extrapolations we developed and validated pedotransfer functions (PTFs) to replace the missing values. The BD for STATSGO and SSURGO were assigned using PTFs for consistency. Total soil water holding capacity, water retention at various matric potentials, and hydraulic conductivity were estimated using existing PTFs that we validated. Sharp differences across detailed survey boundaries were minimized by merging map unit component property databases into regional datasets. Regional soil properties are estimated by using STATSGO mapunits overlayed on data derived from SSURGO and the NSCD. Error analyses of STATSGO/NSCD-derived properties are being made using SSURGO/NSCD-based estimates. The spatial patterns of soil C, N, soil water properties along with an evaluation of the large influences of soil bulk density, texture, rock fragment content, and soil depth will be presented. Portions of the coastal fog belt and coastal mountains had the greatest C, N, and water retention followed by the Willamette Valley because of deep soils frequently low in rock content. The Oregon Cascade Mountains had deeper soils than previously predicted, but rock fragment content decreased the otherwise high C, N, and water retention.