Forest soils and topography have long been known to influence forest productivity in complex terrain such as Oregon’s Coast Ranges. Incorporating physical site characteristics into predictions of forest growth and yield, however, has been problematic because of the high spatial variability of soil properties and the challenges associated with representing topography and soil properties at the stand level. We assessed the efficacy of using a combination of a spatially intensive surface soil sample and terrain indices derived from a digital elevation model for explaining local variability in forest productivity across a forested watershed in northwestern Oregon. Analysis of correlations between several textural characteristics of the surface soil horizon and the average values for the top 100 cm showed that surface soil samples could provide an index of deeper-soil water holding capacity. Analysis of a time series of soil moisture observations from 34 locations in the Panther Creek Watershed revealed that surface soil texture, topographic wetness index, and slope can be used to predict relative soil moisture availability near the end of the growing season.We integrated the findings of the first two analyses to assess the influence of sur- face soil texture, hydrology, and topography on fine-scale variability of basal area periodic annual increment, height increment, and site index across five intensively- managed Douglas-fir plantations within the Panther Creek Watershed. Basal area periodic annual increment was maximized on sites that slope approximately to the south-southwest on the shoulder of ridges, and did not appear to depend on any surface soil characteristics. In contrast, site index was maximized in north-facing draws and increased with clay content of the surface soil.