The equilibrium phosphate potentialSP and the slope of the quantity/intensity (Q/I) relation of three contrasted soils increased as the soil/solution ratio was decreased from 115 to 1/50. On allowing the soil phosphate to achieve equilibrium by prolonged storage under constant environmental conditions, the equilibrium potentialSP and the slope dQ/dI remained constant and independent of the soil/solution ratio. The 'soil/solution ratio effect' is interpreted in terms of phosphate disequilibrium, a condition common in soils in situ, which can be due to either non-uniform removal of phosphate or to non-uniform distribution of soluble phosphate recently added to the soil. The two kinds of disequilibrium cannot be distinguished by their observed soil/solution ratio effects, but may be identified by the direction of change of the potentialsp on the attainment of equilibrium in the soil. PotentialSP values measured in initially phosphate-free solutions are discussed with reference to their dependence on the soil/solution ratio. Aslyng's method of extrapolating to 'zero dilution' gives a reasonably accurate measure of the true potentialSP of an 'equilibrium' soil, but is in error for a 'disequilibrium' soil. It is suggested that the equilibrium potentialSP determined at a 1/50 (or smaller) soil/solution ratio will provide an adequate measure of the mean potentialSP of a disequilibrium soil.