Phosphorus (P), an essential nutrient for plant growth and animal needs, has been identified as an ubiquitous water quality impairment in the United States. In Kansas, a major agricultural state, P loading to the surface waters is a top priority because of the critical role of P enrichment in eutrophication processes and resultant water quality degradation. The objective of this study was to quantify the P sorption and desorption of both stream sediments and upland soils in two Kansas watersheds with contrasting degrees of animal agriculture; Upper West Emma Creek (UWEC) and Red Rock Creek (RRC) watersheds. In-stream sediments were collected from banks, pools, riffles and depositional features. Soils were sampled from wheat, row crop, pasture, and manure-amended fields. Stream water samples were taken under baseflow and storm flow conditions. Our analyses of sediments and soils included equilibrium P concentration at zero net P sorption (EPC[subscript]0), maximum adsorption capacity (P[subscript]max), anion exchange extractable P (P[subscript]lab) and degree of P saturation (DPS). Water samples were analyzed for dissolved reactive phosphorous (DRP). Bank erosion pins were installed in order to estimate bank erosion rates in both watersheds. Results showed that in-stream sediments do not have much more sorption capacity remaining indicated by low P[subscript]max and high DPS. A comparison between mean P[subscript]lab of stream sediments (8.8 mg P kg[superscript]-1 soil) versus field soils (61.2 mg P kg[superscript]-1 soil) reflected that they represent a relatively minor long-term P supply. Of the stream sediments, bank soils had the highest Plab concentrations (24.8 mg P kg[superscript]-1 soil) and would be the largest in-stream P source. Manure-amended fields had the highest Plab (118.6 mg P kg[superscript]-1 soil) due to continued inputs of manure-based P; therefore, representing a large available P pool. Bank erosion contributed about 41% and 11% of the total sediment load in UWEC and RRC respectively. Sediments loads indicated that RRC has more upland sediment inputs than UWEC. Moreover, DRP during storm flow was higher at RRC than UWEC, indicating higher P inputs in RRC from upland soils. Finally, in order to minimize P inputs to the stream system, bank stabilization should be addressed in UWEC and upland best management practices should be implemented in RRC.