The Phosphorus Index (P Index) for Texas was adopted in 1999. Very little validation or evaluation of the model had been done when it was put into use. Research was conducted in the cattle feedlot and dairy areas of Texas to determine the validity of the P Index and recommend changes to improve the predictive ability of the model. The P Index is an 8 X 5 matrix that yields a rating related to the potential for P to be transported from fields. The matrix consists of source and transport factors and their potential to contribute to runoff P. The source factors are soil test P rating, fertilizer P application rate, organic P application rate, fertilizer P application method and timing, and organic P application method and timing. The transport factors are proximity of nearest application area field edge to named stream or lake, runoff class (includes slope, soil vegetation coverage, soil hydrologic condition, soil hydrologic group and runoff curve number) and soil erosion from all sources. The individual site characteristic factors for each of the source and transport factors are multiplied by a weighting factor from 0 to 8 with ratings of very low (0), low (1), medium (2), high (4), and very high (8). These values are added together to give a P Index runoff potential rating. This rating (very low-low, medium, high, very high) is then used in the US Department of Agriculture (USDA)/Natural Resources Conservation Service (NRCS) nutrient management practice standard (Code 590) for Texas to give guidance on the amount of P that should be applied to fields, especially from animal manure, biosolids and other organic sources of nutrients. Eight sites from the cattle feedlot and 40 sites from the dairy areas of Texas were selected to evaluate the P Index. Rainfall simulations following SERA-17 protocol were conducted on each of the sites using a simulated rainfall rate of 7.5 cm/hr. Runoff water samples were collected at 15 and 30 minutes after runoff was initiated and a composite. Volume of runoff was recorded every minute after runoff initiation. Runoff samples were analyzed for pH, EC, nitrate-N, P, K, Ca, Mg, Na and S. Soil samples were collected after the third rainfall simulation at depths of 0-7.5, 7.5-15 and 0-15 cm. These samples were analyzed for pH, EC, nitrate-N and Mehlich-3 P, K, Ca, Mg, K, Na and S. Additional P extractants included Bray I, Olsen, TAMU, 0.01M CaCl2, 0.5 and 0.1M KCl, and DI water and other extractants to estimate inorganic and organic P and P fractions. Other soil extactants were ammonium oxalate for Fe, Al and Si, determination of % carbonates and organic C. Results indicate that by adding Mehlich-3 extractable Mg to the P Index the predictability of soluble P and total P in the runoff increases from an r2 of 0.6 to 0.8. It was found that the Mehlich-3 was a good predictor of total inorganic P and total P in calcareous soils. It was a better predictor in calcareous soils than in acid soils. The parameter that best predicted the soluble and total P in runoff water from rainfall simulations was the 0.1M KCl extractant with an r2 of 0.8.