Background: Soil spatial variability in vineyards can result in differences in yield and quality that cannot be attributed to differences in climatic conditions between growing seasons. Understanding soil variation in vineyards is therefore an essential component of precision viticulture. Materials and Methods: We assessed soil variability in a 4.5 hectare Napa Valley Pinot Noir vineyard. The pedological observations were made in relation with nearby vines that are being monitored for vegetative vigor, yield and fruit quality over a period of several years. Vine vigor was assessed by measuring vine trunk diameters and pruning weights, and by examining NDVI images. The study site was set on a complex convexo-concave slope cutting across a thick series of Pliocene estuarine river alluvium with contrasted granulometry. Soil types include Haploxeralfs, Xerorthents and Haploxererts, depending on underlying parent material, position on the slope, and recent cultivation practices. A preliminary survey included 44 auger holes regularly spaced on the slope, divided into 6 toposequences. Subsequently, 12 soil pits, forming 3 toposequences, were opened, described and sampled. The large number of soil sampling points was selected in a pattern determined by the vines from which data are being taken, and approximating a grid. A primary goal was to determine the smallest number of sampling sites necessary to obtain a realistic large-scale map of soil variability. The soil samples were analyzed for soil texture, CEC, organic matter content and exchangeable cations. This information was then combined in a geodatabase with field observations such as coarse fragment content, soil color and horizon thickness and analyzed using geostatistical methods. Lateral continuity of horizons and the presence of abrupt thresholds of change were examined. The optimal sampling density to conserve the greatest amount of information on soil characteristics was estimated by varying the number of observations included in the study. Results and Conclusions: Soil physical and chemical properties indicated a very heterogeneous vineyard. The study site showed great parent material variability, which typically requires a larger than expected number of sampling points to allow precise mapping. Cultivation of the vineyard has masked natural variability in the surface horizon. Furthermore, initial grading was performed on the slope prior to planting 12 years ago and modified soil classification by truncating or burying natural profiles, leading to irregular C decrease with depth. A high degree of variation between adjacent areas within the vineyard block complicates attempts at site-specific management. Based on the K/CEC ratio, soils in the vineyard had good base saturation but insufficient available K, except in the surface horizon, as a result of fertilization. Vine vigor was lower in areas of coarser soil surface and subsurface texture and associated low water availability. In other areas, factors such as slope, drainage, and amount of backfill prior to planting must be considered in order to explain differences in vigor. The use of GPS to record plant and soil information in a GIS database, along with the use of geostatistical tools, has given a new impetus to vineyard classification and terroir studies by providing a means of visualizing complex soil properties and modeling the soil-vine relationship. The knowledge gained from detailed soil studies offers an opportunity to implement different vineyard management strategies.