Eutrophication due to Phosphorus (P) is a major problem around the world due to runoff from dairy pastures. North central Texas is one such area where most of its watershed is being classified as impaired by Environmental Protection Agency (EPA), due to P contamination of surface waters. This area houses a large number of Confined Animal Feeding Operation (CAFO) dairies with historic manure application and have been recently subject to strict monitoring for P management. Soil conditions and management practices vary greatly among dairies which make a difference in runoff potential of soil P and Phosphorus Index (PI) calculations. The present study was initiated to study the spatial variability of soil test P that exists across dairies with historic manure application. Two dairies were selected in Erath county of north central Texas which are located on the Bosque and Leon river watersheds. The dairies were divided into hypothetical grids of one acre area with four sample points equally spaced around the grid. Surface soil samples (0-12 cm) were collected from each sample point and composited into one sample per acre. A Garmin GPS unit was used to locate geo-referenced sample points on the dairies. Soil test P was analyzed in the samples using Mehlich III extractant. Environmental P, also called runoff P was measured using deionized water as the extractant. ICP-AES was used to estimate the P in both the methods. Other soil parameters like organic carbon percentage, calcium carbonate percentage, pH, EC and clay percentage were also measured. ArcGIS 9.1 software (ESRI) was used for interpolating maps and geo-referencing. High levels of soil test phosphorus were observed and the values ranged from 43 ppm to 3216 ppm. Water soluble P ranged from 1.8 ppm to 312 ppm. A soil test P map clearly depicted higher P levels around effluent application areas and barns. The levels were also higher around lagoons and along drainage canals which ran across fields. Both water soluble P and soil test P were found to be higher in depressions. Lower P levels were observed in vegetative strips along edges of creeks. A significant positive correlation was observed between soil test P and slope. P levels also varied with soil types and management practices. Pasture soil extracted more water soluble P than un-cropped areas. Strong correlations between Mehlich III, clay content, and pH were observed, whereas weak correlations between Mehlich III and calcium carbonate were observed. Weak correlations were observed between water soluble P and other soil properties. Mehlich III was more stable and extracted higher P levels under varying soil conditions, however water soluble P varied based on soil conditions. Thus, the question arises whether PI calculations must be based on soil test P or water soluble P which represents actual amount of P susceptible to runoff. The interpolated maps were developed for soil test phosphorus with varying number of sample sizes and there was a considerable variation among them. Breakpoint in sample size for a true representation of the field found to vary from dairy to dairy depending on the soil type and management practices. There were many factors which affected P buildup and water soluble P. Thus, questions concerning current sampling intensity, testing methods, and PI calculation procedures deserve further attention and research.