Sediment toxicity tests are ecotoxicological tools for the environmental risk assessment and have increasing importance for regulators, scientists and technologists. Natural or formulated sediments are used in bioassays and the nature of the substrate may influence the results. The purpose of this study was to test the influence of sediment type on the toxicity of polycyclic aromatic hydrocarbons (PAHs), ubiquitous contaminants of natural sediments. The toxicity of sediment contaminated by phenanthrene only or a PAHs mixture (phenanthrene, fluoranthene, benzo(k)fluoranthene) were assessed using single and multi-species test system on the aquatic invertebrates, Hyalella azteca and Chironomus riparius. Sediment tests on these two freshwater organisms are recommended because of their sensitivity, ecological relevance, amenability, sediment tolerance and availability of methods. Moreover the influence of sediment type on the toxic effect of PAHs was studied using 3 types of substrates: two natural sediments, one formulated sediment and one formulated and conditioned sediment. The influence on PAHs toxicity of a 8-d equilibration step of experimental systems before test initiation was also studied in multi-species tests involving PAHs mixture. Fluoranthene and phenanthrene were chosen for their abundance in sediments, and the benzo(k)fluoranthene as regards to the lack of toxicity data for this substance. Their structure and solubility are different : 3, 4 and 5 aromatic rings, a solubility limit in water of 1000, 206 and 1.5 µg/l and, a log Kow of 4.57; 5.23 and 6.11, respectively for phenanthrene, fluoranthene and benzo(k)fluoranthene. The mono-substance exposure (phenanthrene in the present study) represents the usual practice to determine SQVs and the mixture exposure allowed to improve the relevance of risk assessment with reference to environmental pollutions. The fate and effects of natural and formulated freshwater sediments-associated PAHs (phenanthrene or PAHs mixture) were assessed in single- and multi-species tests over 10 or 14 days. Clear influence of sediment type on the PAHs partition and toxicity was observed. NOEC were ³ 300 mg/kg of phenanthrene and ³ 30 mg/kg of PAHs mixture in natural sediments whereas LOEC in formulated sediment were 30 mg/kg of phenanthrene and, in most cases, £ 18 mg/kg of PAHs mixture. Differences of toxicity seemed to be only partially related to the OC content. Our results indicated that other factors than OC quantity (e.g. organic matter nature, inorganic sediment composition) influences organic compound toxicity and bioavailability. The formulated sediment used in this study induced higher toxic effects than natural sediments and seemed to simulate poorly the natural sediment properties. Test conditions, such as formulated sediment conditioning or short-term equilibrium of experimental system before test initiation, influence the toxicity results and need standardisation in order to allow relevant comparison between inter-laboratory studies.