The objective of this study was to test a simple input‐output model to estimate soil phosphorus (P) retention over 12 years in a Mediterranean pasture. A 6‐ha permanent pasture field in Southern Portugal was subjected to different management systems in terms of grazing and fertilizer application. The soil and pasture were monitored and samples were geo‐referenced with RTK GPS. The model was developed to explain the effect of P fertilizer, based on the premise that P added through fertilizer and amounts removed in harvested plant products are the main fluxes affecting available P status, and was used to estimate the soil P retention. The results show that P fertilizer addition to Mediterranean grazed pastures on shallow soils influences several P dynamics under different grassland management systems. A positive linear relationship between the P budgets and the changes in measured P was demonstrated over the study periods. This work highlights the difficulty associated with the creation of recommendation maps for differentiated P fertilizer application in pastoral systems and indicates that before any accurate decision making, the historic knowledge of changes in the concentrations of P in the soil should be considered, as well as a prediction of the residual effects of applied nutrient. The importance of this knowledge increases when the aim is to use techniques and technologies for variable‐rate fertilizer application and implementation of site‐specific management strategies in the context of precision agriculture.