Phosphate fertilizers are used all over the world for increasing the fertility of agricultural land. Apart from organic P fertilizers, mineral P fertilizers are most commonly used. They contain between below 1 and more than 100 mg U/ kg. After application and entering the soil zone, this toxic element might finally reach groundwater. Since November 2011, after long discussion, an MCL of 10 μg U/ L was included in the German Drinking Water Directive. However, until now there are no regulations on how much U is allowed to be in phosphate fertilizers. There is not even an obligation of indicating the U contents of commercially available P fertilizers. Many studies do exist on the sorption of U on different pure or mixed substances and also some long-term field studies. However, there are still numerous unknowns on the fate of U in the subsurface and processes and factors, which control that. So, the main goals of this thesis were to compare different agriculturally used sites with corresponding forest sites by interpreting available long-term soil-, percolation water and groundwater data, to take own soil and water samples to characterize their composition and properties, to conduct various tests with them, to model the speciation in the water samples, and to finally put all these findings together. For characterizing soils' properties, grain size distribution, pH, carbonate and organic carbon content as well as cation exchange capacity were determined. Distilled water, ammonium acetate at pH 7 and 5, EDTA, hydroxylamine hydrochloride (NH2OH*HCl) and 7 N nitric acid were assessed for their U extraction capability from soils. It was examined whether the specific activity ratio of U-238/Ra-226 can be used for elucidating the origin of U (geogenic, i.e. from P-containing rock, or anthropogenic, i.e. from P fertilizers). From three sites, water samples were taken and analyzed for main and trace elements using IC and ICP-MS, respectively. The results were used for modeling U and P speciation and saturation indices using PHREEQC. Flow-through experiments under saturated conditions using field top-and subsoil samples from one of the study sites, the lysimeter station in Brandis, were conducted using solutions containing 1 mM NaHCO3, 0.168 μM UO2(NO3)2, no, 1. 421 or 0.142 mM H3PO4 and some of them also 1 mM Ca(NO3). The applied P concentration was based on the GAP (good agricultural practice) guidelines, and U concentrations were selected according to typical mean contents in P fertilizers. By increasing the contents applied per time, a time shift of 319 days was achieved and consequently a time of 93 years could be performed within 106 days. From field studies it was found that no general conclusion on the fate of U can be made. Contents in soils varied between 1.3 and 4.1 mg/kg; contents in waters between below detection limit (0.001 μg/L) and 12.9 μg/L. At some sites, concentrations remained constant over time, at others, they increased or decreased. Soil and groundwater U contents did not show the same trend at each study site. The same applies for U contents in groundwater of monitoring wells in the surrounding. The behavior of U depends on soil properties and composition, as well as fertilization (applied amounts, type of fertilizer, U content, season and frequency of application), climate (e.g. precipitation), ploughing practices (frequency, depth), and others. Distilled water, ammonium acetate (pH 7), EDTA and hydroxylamine hydrochloride mobilize below or around 1% of total U; ammonium actetate (pH 5) set free between about 2 and 9%. Even when nitric acid was used, extracted amounts were only between 15 and 40%. This suggests that U is firmly bound as other studies confirmed. Repeating the ammonium acetate (pH 5) step with the samples from flow-through experiments, recoveries were 5-to 6-times higher. This supports the findings of other scientists that U binding gets stronger over time. Low-level gamma spectrometry could not serve to reveal the origin of uranium because uncertainties were bigger than differences etween field and corresponding forest soils. During flow-through experiments, about 99% of total U was retained even if species modeling suggested uncharged, mobile U complexes to account for a great or the greatest share of total U in most flow-through cells. So, comparability with findings from the field is very limited. The reasons for the effective U retention during flow-through experiments is not yet clear but changing redox conditions in micropore regions of soils, the action of microorganisms and the formation of solubility-controlling phases might play an important role.