Phosphorus (P) loss from land can impair surface water quality. Losses can occur from soil and plant components. While it is known that P losses increase with soil P concentration, it is not known how losses from pasture plants vary with soil P concentration or between different forages. We examined total P and filterable reactive P (FRP) in water extracts of plant shoots, used as a measure of potential P loss to surface runoff, in different forage species relative to soil P concentration in field trials and a glasshouse experiment. The mean total P concentration of 16 forage species in grazed field plots was greater (P<0.01; LSD 05 =117 mg kg −1 ) in legumes (3,480 mg kg −1 ) than for grasses (3,210 mg kg −1 ). Total plant P concentrations of grasses and legumes increased with soil Mehlich-3 P concentrations in both glasshouse and field trials with concentrations close to 6,000 mg kg −1 in arrowleaf clover at 680 mg kg −1 Mehlich-3 soil P. FRP in water extracts of plant shoots increased relative to plant total P as soil Mehlich-3 P increased, with the greatest concentrations shown by crimson clover and arrowleaf clover. Analysis of water extracts of ryegrass and clover herbage from a field trial showed that while FRP was increasing, phytase-available-P decreased significantly from about 70% of filterable unreactive P at the lowest Mehlich-3 P concentrations, to close to zero at 200 mg kg −1 Mehlich-3 P. The wide variation, and enrichment of FRP in water extracts and total P with increasing Mehlich-3 P among species, indicates that cultivar and site selection and sward management provide a potential option to mitigate P loss to surface waters.