We analyzed the linearity of relationships between soil test P (STP) and cumulative phosphorus (P) budget using data from six long-term fertilized grassland sites in four countries: France (Erce´ and Gramond), Switzerland (Les Verrie`res), Canada (Le´vis), and Finland (Maaninka and Siikajoki). STP was determined according to existing national guidelines. A linear-plateau model was used to determine the presence of deflection points in the relationships. Deflection points with (x, y) coordinates were observed everywhere but Maaninka. Above the deflection point, a significant linear relationship was obtained (0.33\r 2 \0.72) at four sites, while below the deflection point, the relationship was not significant, with a negligible rate of STP decrease. The relationship was not linear over the range of STP encountered at most sites, suggesting a need for caution when using the P budget approach to predict STP changes in grasslands, particularly in situations of very low P fertilization. Our study provides insights and description of a tool to improve global P strategies aimed at maintaining STP at levels adequate for grassland production while reducing the risk of P pollution of water.