Soil Phosphorus (P) leaves agricultural fields dissolved in runoff water and bound to the sediment eroded from the field. Phosphorus source may affect the loss and/or partitioning of P between sediment and runoff. Limited research exists on phosphorus sources and their potential for runoff from alkaline, calcareous soils. Two related research projects were conducted by Colorado State University to determine P runoff behavior on calcareous soils. The objectives of these experiments were to determine (1) the impact of soil CaCO3 levels on the relationship between soil test P (Mehlich 3) and runoff P levels and (2) P source influences on runoff P concentrations. Experiment 1 took place near Sidney, Nebraska on a fine-loamy, mixed, mesic Aridic Haplustoll with 1% calcite by weight; Tribune, Kansas on a fine-silty, mixed, mesic Aridic Haplustoll with 4% calcite by weight; and Fort Collins, Colorado on a fine-loamy, mixed (calcareous), mesic Ustic Torriorthents with 9% calcite by weight. Eight treatments were replicated at each location: 0, 11, 22, 45, 90, 135, 202, and 269 Mg beef manure ha-1. Experiment 2 took place near Fort Collins, Colorado on a fine-loamy, mixed (calcareous), mesic Ustic Torriorthent with 9% calcite by weight. Four treatments were applied at two rates (90 kg P2O5 ha-1 and 180 kg P2O5 ha-1): raw dairy manure, windrow compost, vermicompost, and rock phosphate. All manure-based treatments were made from the same dairy manure. The same field methods were used for both experiments. Runoff plots were constructed measuring 2 m long and 1.5 m wide. The rainfall rate was 8.3 cm hr-1. After runoff began, 1 L of runoff was collected at 2.5-minute intervals for a 30 minute period. Samples were analyzed individually for their chemical content, runoff volume and sediment mass. Soil samples (0-15 cm) were collected prior to completion of rainfall simulation. Water and soil samples were analyzed in the laboratory using the same methods for both experiments. Filtered (.45 µm), unacidified water samples were analyzed for dissolved inorganic P using the ascorbic acid colorimetric method of Murphy and Riley (1962). Filtered and acidified (with HCl to pH 2) water samples were analyzed for total dissolved P using inductively coupled plasma spectroscopy (ICP). The unfiltered, acidified water samples were measured for total P after digesting with HClO4 and HNO3 (Watanabe and Olsen, 1965) on an ICP. Olsen P extractions were done with 0.50 M NaHCO3 (Olsen et al., 1954), and the P concentration was determined colorimetrically using the ascorbic acid method (Murphy and Riley, 1962). Mehlich-3 P soil extractions followed the method outlined by Mehlich (1984) and were measured by ICP. In Experiment 1 the Kim soil with the highest calcium carbonate concentration sorbed the most phosphorus; however, this resulted in the lowest percentage of P saturation in the soil due to the high Ca concentration. A multiple regression equation was developed across locations to predict total dissolved P in runoff from soil test P and percent CaCO3 by weight. The R2-value is 0.92 using the Mehlich-3 extractant and 0.81 using Olsen's. Experiment 2 concluded that Mehlich-3 predicted both total and dissolved P runoff concentration on a calcareous Colorado soil while Olsen P best forecasted total dissolved P on a Kim soil. Also, the addition of raw manure resulted in greater concentrations of P in runoff than the composted and vermicomposted manures. Therefore, the source of P added to an agricultural field does affect the potential for the P to leave the field dissolved in runoff or attached to eroding soil particles. In conclusion, to minimize P runoff from land applied manure, farmers will be encouraged to compost manure prior to application or to select fields with higher CaCO3 levels to receive manure.