The objectives of the present study were: (1) to evaluate the predicting value of the most important European soil P tests for P leaching losses; and (2) to investigate how these soil P tests reflect the development of P depth profiles in original homogeneous soils of lysimeters. The study included more than 100 lysimeters, located at the Lysimeter Station Falkenberg/Saxony-Anhalt, UFZ-Centre for Environmental Research Leipzig-Halle GmbH, Germany. Soil textures were sand, sandy loam, loam and silt. The management forms were arable land, grassland and fallow with various variation in fertilisation, crop rotation and irrigation. Samples were collected from the A-horizons and from the whole profiles of eight set-aside and dismantled lysimeters at 10-cm sections. The concentrations of total P were determined monthly in the leachates and evaluated for a three-year period. The concentrations of P extracted by ammonium acetate lactate (AL-P), double lactate (DL-P), sodium bicarbonate (Olsen-P) and ammonium oxalate (OX-P) as well as P_t were significantly correlated with each other (P<0.05–P<0.001) for arable soils. The relevant regression coefficients were strongly influenced by soil texture, soil use and management. The mean annual P concentrations of the leachates were in the range 0.4–1.2 mg l^-1 for sands and <0.001–0.1 mg l^-1 for the textures sandy loam, loam and silt. These corresponded to P leaching losses of 0.001–2846 g ha^-1 yr^-1. Mean annual and maximum P concentrations and leaching losses were significantly (r>0.954, P<0.001) predicted by the OX-P concentrations of arable topsoils in lysimeters filled with sand. For sandy loam under grass the agronomic soil P tests (AL-P, DL-P and Olsen-P) enabled reasonable predictions of P in leachate. Under arable use, factors such as fertilisation, management intensity, depth of tillage and irrigation resulted in weak correlations between soil P concentrations and P in leachate. It was shown for the first time that all P extractants reflected P enrichments in topsoils and subsoils and the development of distinct depth profiles. Influence of soil use on the depth distribution of P was more pronounced in the 0–20 cm layer than in the subsoils. Here, the original homogeneous substrate had oscillating P concentrations at 10-cm increments under all soil uses. These could not be explained by Al_ox and Fe_ox but were significantly correlated with the C_t contents and bulk density. This indicates that vertical movement of P containing organic matter along with differences in porosity contributed to the heterogeneous P distribution in the lysimeter subsoils. This new evidence must be considered if data sets from long-term lysimeter experiments are used to calibrate and validate P leaching models.