Predicting effects of chemicals in real life situations is a very complex undertaking, but also very important in terms of the sustainability of our environment. The complexity lies in the enormous variation that is found in real life situations. A variety of organisms is exposed to a variety of chemicals, with a different composition in space and time. This is far from a laboratory situation, where single organisms are exposed to mostly single compounds in a homogeneous environment.As 20 different compounds already allow more than one million possible mixtures we need tools in the form of mathematical models to assess real life exposure situations. These tools must be able to extrapolate from one compound to a mixture containing that compound, from one compound to a compound that was not measured, from effects at some point in time to different points in time and from one species to another.The first steps in the development of such a model were taken during my PhD research, we showed that it is possible to predict effects on a single species, exposed to a real life mixture of metals, pesticides, PAHs, PCBs, nutrients, salts, etc. with great accuracy.The aim of the proposed research is to extend the current single-species approach to a multi-species approach. This would be an important step forward as such an approach is of great practical importance in environmental quality management, understanding and predicting effects of chemical pollution on biodiversity and in understanding and predicting environmental effects caused by accidents and spills. We will build on the existing model, further develop the model to predict effects for untested species, apply QSAR approaches to predict effects for untested compounds and make use of existing experimental data gathered by the UK Environmental Agency on exposures, effects and the occurrence of species to evaluate the approach.