Phosphorus (P) released from plants after freezing at the end of the growing season may be a contributor to P in runoff from agricultural landscapes. We evaluated P release from alfalfa (Medicago sativa L.) and mixed grass species [mainly quackgrass (Agropyron repens)] after freezing or drying in laboratory and field studies. Freezing released 18-30 % of the total P as dissolved reactive P (DRP) and 11-25 % of the total P as soluble P. Drying released more water soluble P than freezing with 30 and 74% of the total plant P released as DRP and total soluble P respectively. Freezing plants in the laboratory or infield treatment with the herbicide paraquat (1,1’-dimethyl-4, 4’-bipyridinium ion) greatly increased water-extractable P, with more P extracted from grasses than from alfalfa. Alfalfa grown on soils with excessive P soil test levels released more P after freezing than plants grown on soils with optimum P levels. A strong correlation (R2 = 0.97) was found between soluble P released by freezing and total P in alfalfa collected from fields with soil test P levels ranging from 35 to179 ppm. Runoff from paraquat-treated alfalfa or grass field plots subjected to simulated rainfall contained higher P concentrations than runoff from untreated plants. The effects of natural freezing of alfalfa on P losses in runoff were evaluated by collecting runoff from alfalfa and control plots during the October through May over winter period. Although laboratory and simulated rainfall experiments showed the potential for P losses after plant freezing, alfalfa treatments did not affect P or sediment loads in natural runoff in one year of measurement. Climatic conditions including the timing and extent of plant freezing and drying and of precipitation events after freezing likely influence the potential for P losses in natural runoff.