Cuttings of Populus nigra L. cv. Loenen were cultivated in sand treated with one of the following PAHs: phenanthrene (Phen), fluoranthene (Flt), pyrene (Pyr), chrysene (Chr) and benzo[a]pyrene (BaP). The PAHs were applied at varying levels of concentration to each test series. After 6 weeks the concentration and the distribution of the PAHs in the substrate of the various sets of tests were compared with the concentration in the substrate of the control. Additionally the substrate and the plant roots were tested for evidence of degradation products of PAHs. The results revealed that the levels of concentration of Phen and Pyr detected in the substrate surrounding the roots was in some cases significantly lower than in the corresponding section of substrate in the unplanted set (= control). This phenomenon did not occur for Flt and BaP and in the case of Chr only in those substrates, which had been treated with the highest levels of concentration. As the presence of lesser amounts of Phen and Pyr in the plant pots cannot only be attributed to their accumulation and metabolism in the roots, it is fair to assume that the chemical transformation of these three PAHs took place outside the roots. The set of tests treated with Phen revealed the presence of 2- or 3-hydroxy-Phen (main components), a hydroxy-methoxy-Phen, 9,10-Phenanthrenequinone and one unidentified compound in metabolite form. Altogether eleven metabolites of Pyr were identified in the root extracts, which can be divided into three groups: 1-Hydroxy-Pyr and derivatives, dihydroxy-Pyr and derivatives and ring fission products (4-Hydroxy-Pyr and a derivative of the 4-Phen-carbonic acid). However, the metabolite mass detected for Phen and Pyr represents only an insignificant percentage in comparison with the lesser amounts of PAHs observed in the planted set of tests. This indicates that the three PAHs were reduced to lower molecular compounds, which are methodically impossible to record, and subsequently translocated to other parts of the plant and integrated into the biomass. Although no lesser amount for Flt and BaP was found in the plant pots, 1-Hydroxy-Flt, an unidentified compound of Flt and 1-Methoxy-BaP were detected. These are presumably end products which were enhanced in the roots. It was not possible to identify any transformation products of Chr. It can be assumed that the majority of metabolites were not synthesised in the roots but are a result of microbial degradation in the rhizosphere. The test plants improved the conditions for the biotransformation of Phen and Pyr significantly and accumulated Flt, Pyr, Chr and BaP in their roots. It can therefore be concluded that the use of plants in the bioremediation of contaminated soils is a promising option.