The U.S. Environmental Protection Agency's watershed approach to address impacts of nonpoint source pollution on water quality uses the Total Maximum Daily Load (TMDL) concept. An important component of TMDL development is watershed characterization, which encompasses loadings, transport processes, and in-stream levels. The objective of this study was to assess phosphorus (P) loadings for various sources using soil P, soil properties and land use in an agricultural watershed. The Pontotoc Ridge/Flatwoods Branch Experiment Station, which encompasses portions of Town Creek, Tibbee River, and Yalobusha watersheds, was chosen as the study site because of the availability of historical information on land use, management practices, and data relevant to the calculation of P loading. The study site lies on two Major Land Resource Areas (MLRAs): the Upper Coastal Plain MLRA, comprised of sands, clays, and shale, and dominated by alfisols and ultisols, lies on the eastern side of the station; and the Interior Flatwoods MLRA, which is underlain by gray acidic shale of poorly drained alfisols and entisols. It had a diverse land use distribution of agronomic crops (corn, cotton, soybeans, forage, and sweet potatoes), horticulture, and timber and swine production. Grid samples were collected to evaluate the landscape distribution of P in relation to land use, slope, drainage, and other classification groupings, and to calculate P loadings. Soil samples were collected from 0- to 15-cm depth at the center-point of 0.6 sq ha for a total of 400 samples. These were analyzed for Mehlich III, Olsen, and total P. Mehlich III-P values ranged from 0.6 to 109 mg P kg soil-1, Olsen-P from 46 to 216 mg P kg soil-1, and total-P ranged from 90 to 668 mg P kg soil-1. Land use and geology were the major influences on P distribution. Swine production waste management was a dominant influence on P levels, regardless of soil test method. The Interior Flatwoods MLRA P loadings for were greater than those for the Coastal Plain. Forest land use had notably high levels of total P. The <8% and 8 to 12% slope classes had significantly higher mean Mehlich III- and Olsen-P than other slope classes. Soils with 9-18 percent clay had significantly higher Mehlich III-P concentrations than soils with 18-27 and 27-40 percent clay. High levels of soil-P are not necessarily equivalent to a high level of environmentally mobile P. Phosphorus distributions in agricultural watersheds should be related to runoff and in-stream P concentrations in order to appropriately manage the contribution of these sources.