Phosphorus (P) exports from agricultural land are a problem world-wide and soil tests are often used to identify high risk areas. A recent study investigated changes in soil (0–20 mm), soil water and overland flow in 4 recently laser-graded (<1 year) and 4 established (laser-graded >10 years) irrigated pastures in south-eastern Australia before and after 3 years of irrigated dairy production. We use the results from that study to briefly examine the relationships between a series of ‘agronomic’ (Olsen P, Colwell P), environmental (water-extractable P, calcium chloride extractable P, P sorption saturation, and P sorption), and other (total P, organic P) soil P tests.

Of the 2 ‘agronomic’ soil P tests, Colwell P explained 91% of the variation in Olsen P, and Colwell P was better correlated with the other soil tests. With the exception of P sorption, all soil P tests explained 57% or more of the total variation in Colwell P, while they explained 61% or less of Olsen P possibly due to the importance of organic P in this soil. Variations in total P were best explained by the organic P (85%), Calcium chloride extractable P (83%), water-extractable P (78%), and P sorption saturation (76%). None of the tests adequately predicted the variation in P sorption at 5 mg P/L equilibrating solution concentration.

The results of this limited study highlight the variability between soil P tests that may be used to estimate P loss potential. Moreover, these results suggest that empirical relationships between specific soil P tests and P export potential will have limited resolution where different soil tests are used, as the errors in the relationship between soil test P and P loss potential are compounded by between test variation. We conclude that broader study is needed to determine the relationships between soil P tests for Australian soils, and based on that study a standard protocol for assessing the potential for P loss should be developed.