In a series of glasshouse experiments, plants were grown in pots and their response to applied phosphate was measured. In the experiments we measured the response of subterranean clover and ryegrass to applied phosphate in surface soil and subsoil, with and without inoculation with vesicular-arbuscular (VA) mycorrhizal fungi, and with different levels of added iron hydroxide. For subterranean clover, there was often a clearly marked threshold level of phosphate application below which the plants took up little P and grew poorly. This threshold effect led to the sigmoidal response curves. It occurred when plants were grown in subsoil and in surface soil when iron hydroxide was added. However, it only occurred when the plants were non-mycorrhizal. Inoculation with a VA mycorrhizal fungi (Glomus fasciculatum) eliminated the threshold even when large amounts of iron hydroxide were present. For ryegrass, no threshold was observed and the response curve was never sigmoidal. Thus sigmoidal response curves to applied phosphate were only observed when a coarse-rooted plant species (subterranean clover) was grown in soils with large adsorption capacities for phosphate, and when roots were not mycorrhizal. Sigmoidal response to applied phosphate may occur because there is a threshold concentration of P in soil solution for adsorption by plant roots, for movement to plant roots or for desorption of adsorbed phosphate from the soil particles.