The capacity of a calcareous desert soil, Thiokol silt loam, to retain natural , as well as added , orthophosphate-P was measured by equilibrium adsorption employing a batch technique and evaluated using the two-slope Langmuir adsorption isotherm. From these data, corrected to account for indigenous soil P, a hypothesis was formulated as to the nature of retention of P by the soil, including the identification of two interfacial reactions involving P, and a value calculated for the adsorption maximum as defined by the Langmuir isotherm equation f o r P with soil at each of two soil depths and t h ree constant temperatures within the range of biological activity . The initial reaction was considered to be surface adsorption , where phosphate ions interact with the clayand lime mineral surfaces at definite sites. The activity of the second mechanism was identified as adsorption, and in addition, the heterogeneous nucleation of metal phosphates on the lime mineral surfaces. In addition to these quantitative studies, the flux of P in the soil was also investigated for the same soil and temperatures by means of kinetic experiments conducted to identify the nature (mechanisms) and measure the rates and release maxima of indigenous P release from the soil. These experiments were carried out using an anion-exchange resin as an infinite sink for P, again applying a batch technique. Ultimately, the release of indigenous P from the soil under saturated conditions was attributed to three simultaneous first-order reactions. 'll1e rate constants of the three reactions were found to be of orders of 10-4, 10-5, and 10-6 (1/sec), and did not vary significantly with soil depth or ternperature. The three reactions above were identified as dissolution of poorly crystalline or amorphous calcium phosphates, the desorption of surface site adsorbed or labile P, and the slow dissolution of calcium hydroxyapatite, respectively.