A phosphorus (P) index for pastures was developed to write nutrient management plans that determine how much P can be applied to a given field. The objectives of this study were to (i) evaluate and compare the P index for pastures, particularly the P source component, and an environmental threshold soil test P level by conducting rainfall simulations on contrasting soils under various management scenarios; and (ii) evaluate the P index for pastures on field-scale watersheds. Poultry litter was applied to 12 small plots on each of six farms based on either an environmental threshold soil test P level or on the P index for pastures, and P runoff was evaluated using rainfall simulators. The P index was also evaluated from two small (0.405 ha) watersheds that had been fertilized annually with poultry litter since 1995. Results from the small plot study showed that soil test P alone was a poor predictor of P concentrations in runoff water following poultry litter applications. The relationship between P in runoff and the amount of soluble P applied was highly significant. Furthermore, P concentrations in runoff from plots with and without litter applications were significantly correlated to P index values. Studies on pastures receiving natural rainfall and annual poultry litter applications indicated that the P index for pastures predicted P loss accurately without calibration (y = 1.16x - 0.23, r(2) = 0.83). These data indicate that the P index for pastures can accurately assess the risk of P loss from fields receiving poultry litter applications in Arkansas and provide a more realistic risk assessment than threshold soil test P levels.