<p>There is broad acknowledgment that soil phosphorus (P) accumulates in agricultural soils over time due to positive P balance where P inputs exceed P outputs in products. Assuming no other constraints, optimum plant yield can be achieved at “critical”, “threshold” or “optimum” soil P levels. Many reviews focus on a single soil test P (STP) for a specified crop, or specific soil types, but this review highlights the variation in critical P levels derived from different STPs and their divergence under varying conditions. In this study, we reviewed more than 100 research papers, published from the mid-twentieth century until 2024. To qualify as a critical P level, the plant, soil test, sampling depth, regression model and relative yield (RY) should be specified. Only 51% of the critical P levels in the cited papers included all 5 of these primary data requirements, while 13% were missing more than one of them. Olsen P was the most used STP for determination of critical P levels and wheat was the most used plant. Based on soil solution and bicarbonate extractants, vegetables and fiber crops ranked highest, indicating high P requirements and less efficient P acquisition. In contrast, legumes and nut crops rank lowest, reflecting lower P needs and more efficient P mobilization and uptake mechanisms. Relative yield targets of 90 and 95% were the most common and provided higher critical P levels compared to lower RY targets. The results show that the Olsen P critical levels increase with decreasing pH and increasing in soil organic carbon. The low level of granularity in the available data suggests that a common approach to critical values and P use efficiency is lacking globally. Adopting evidence-based fertilizer decisions based on the currently available critical levels in each geographic region enables efficient use of P fertilizers, minimizing the risk of P loss from the soil and enhancing profitability for farmers.</p>