Since the United States Department of Agriculture's Conservation Reserve Program was established by the Food Security Act of 1985, thousands of acres of cropland in stream riparian zones have been converted to conservation buffers. The capacity of these buffers to retain phosphorus (P) is related to the P sorbing capacity of the soil in the buffer. Laboratory experiments were conducted to investigate P sorption, desorption and buffering capacity of a long-term buffer and an adjacent crop field. Potential P sorption and binding intensity were evaluated by applying one and two-surface Langmuir models to sorption isotherms from soil collected over three depths, 0 – 5, 0 – 10, and 0 – 15 cm. Overall, the one surface Langmuir model fit the P sorption data well (R² > 0.98). Over the range of equilibrium P concentrations in this study, the ability of the soil to adsorb added P was significantly greater in the crop field than the buffer despite regular P addition to the crop field. Sorption maxima (Smax) ranged from 555 to 625 mg kg^-1 and 417 to 435 mg kg^-1 in the crop field and the buffer, respectively. Binding energy (k) ranged from 0.23 to 0.34 L mg^-1 and 0.16 to 0.24 L mg^-1 in the crop field and the buffer, respectively. Phosphorus equilibrium buffering capacity was significantly lower, and equilibrium P concentration was higher in the buffer. Only the 0-15 cm samples showed a better fit with the two surface model. The fact that the two surface model fit the deeper samples suggests that P binding sites deeper in the soil are different that those near the surface. The study suggests that the effectiveness of buffers intended to reduce P loss in runoff from crop fields depends on the physical and chemical properties of the soil and the management of the buffer