While generally immobile, as a soil becomes saturated with P, the risk of P leaching increases.  Over time, water movement within a P laden landscape should redistribute various P compounds based on their potential mobility in soil, and the properties of the soil across the landscape, creating a unique spatial distribution related to hillslope hydrologic processes.  To test this hypothesis, we collected spatially explicit soil samples on marginal agricultural lands in West Virginia at locations with long histories of N-based manure applications.  These samples were sequentially extracted with H₂O, NaHCO₃, NaOH, and HCl. Molybdate-reactive P and total P was measured in each extract.  Organic P was calculated as the difference between these fractions. Additionally we collected data related to the P sorption capacity (%C, %N, Ca, Al, Fe, pH, etc.) for each sample. We tested the P data for spatial autocorrelation and used a partial Mantel test, to isolate the variability associated with changes in the soil properties from spatial effects. The H₂O extractable fraction and the organic portions of the NaHCO₃ and NaOH fractions were found to be spatially autocorrelated (Morans I values of .21, .25, .46, and .29 respectively). The partial mantel test identified significant spatial effects (p >0.05) on the organic portion of the NaHCO₃ fraction. This spatial effect is assumed to be related to P redistribution.  These results are consistent with the most mobile fractions (H₂O) being extremely transient, the remaining inorganic P and the organic P associated with Fe and Al hydroxides being stable within the landscape, while the moderately labile frsction (the organic portion of the NaHCO₃ fraction) showing evidence of redistribution, possibly from hillslope hydrologic processes