In many areas, continual application of fertilizer and manure has increased soil phosphorus (P) to levels that are of environmental rather than agronomic concern. More information is needed on the long‐term residual availability of soil P to make more reliable P management recommendations. Three Italian soils with a wide range in phosphorus adsorption properties were equilibrated for 91 days with different rates of fertilizer P to obtain similar levels of soil extractable Mehlich 3‐P (M3‐P). After equilibration, P was desorbed from the soils using ten successive extractions with iron oxide‐impregnated paper strips, Mehlich3 extractant, and water solution (Psol). At similar levels of M3‐P contents (57±6 mg P kg‐1soil), more P was released from the soil with higher P sorption index (SI) than from the soil with lower SI, indicating that a greater available P content was necessary from low P sorbing soils to maintain a given rate of P release. Soil M3‐P and Psol decreased in the course of P removal, and a close linear relationship was found between the decrease in soil M3‐P and P released onto the strips (r=0.99; P<0.1%). Furthermore, soil with the highest P buffer power presented the lowest variation in M3‐P and soil P solution values, and changes in M3‐P per unit of P removed were significantly related to soil SI (r=0.97; P<5%). The data suggest that the successive soil P extraction procedure carried out using Fe oxide‐impregnated paper strips may provide a reliable laboratory method for characterizing P desorption from soils and easily estimating residual soil P availability.