A mathematical procedure was developed to permit the estimation of both the amount of phosphate initially displaced from soil and the amount undergoing secondary adsorption. The proportion of previously added phosphate which was initially displaced by bicarbonate decreased as the period and temperature of prior contact between soil and phosphate increased. The proportion displaced was greater than was displaced by dilute calcium chloride. The amount of secondary adsorption depended on the soil:solution ratio, the level of addition of phosphate, the period of shaking, and the kind of soil. Part of the effect of soil:solution ratio appeared to arise because of adsorption of bicarbonate. As a result, at high soil:solution ratios, the bicarbonate concentration in solution was lower and so competition for re-adsorption was less effective. Bicarbonate was less effective at preventing re-adsorption when the level of prior addition of phosphate was high. This may have occurred because prior adsorption of phosphate had increased the negative charge on the surface. Because of secondary adsorption, there was a curvilinear relationship between the amount of phosphate added to the soil and the amount present in the extract.