Samples of soil were collected by colleagues in New Zealand, Thailand, Alabama, California and Hawaii and in four widely spaced locations in Australia. The soils were incubated with a range of levels of added phosphate and the changes in phosphate concentration in the solution were followed for up to 90 days. Eight of the soils were also used in a pot trial which compared the availability of freshly applied phosphate with that of phosphate that had been incubated with the soil. Adsorption was described by a modified Freundlich equation in which the effects of concentration and time were both described by exponent terms. The values of the exponent terms differed between the soils. There was more variation amongst the whole group of soils than amongst the soils collected from within one region. The two exponents were strongly correlated. Differences between the soils in the effects of time, as characterized by the exponent, were not related to the iron or aluminium content of the soil. The soils differed in the extent to which the availability of phosphate to plants decreased when phosphate was incubated with the soil. Equations were formulated to relate the exponent terms measured in the laboratory to the glasshouse measurements of the effect of availability to plants.