A complex chemical cocktail, with unknown composition and concentrations, is present in marine waters. Although the awareness of the vulnerability of marine ecosystems to pollution-induced changes increased, the ecotoxicological effects of chemical pollutants on marine ecosystems are poorly understood. Even in intensively monitored regions such as the North Sea, current knowledge of the ecotoxicological effects of chemicals is limited to few (priority) substances and few (model) species (discussed in chapter I). To partly address this knowledge gap, in the present work, using phytoplankton, it is assessed how marine ecosystems respond to the presence of organic chemicals. By analyzing existing data and performing laboratory experiments, ecotoxicological effects of organic chemicals to marine organisms and ecosystem functions are quantified. Specific aims of this work are: (1) to infer spatiotemporal trends of concentrations of organic chemicals; (2) to investigate the impact of primary and secondary emissions on the spatiotemporal trends of organic chemicals; (3) to examine the partitioning of organic chemicals in different environmental compartments; (4) to assess the potential ecotoxicological effect of realistic mixtures of organic chemicals along environmental gradients; and (5) to quantify the relative contribution of organic chemicals to the phytoplankton growth dynamics. Spatiotemporal trends of polychlorinated biphenyl (PCB) concentrations are inferred based on an extensive set of concentrations monitored between 1991 and 2010 in sediments of the Belgian Coastal Zone (BCZ) and the Western Scheldt estuary in chapter II. The time trends unravel two to three-fold PCB concentration decreases in the BCZ during the last 20 years. In the Western Scheldt estuary, time trends are spatially heterogeneous and not significantly decreasing. These results demonstrate that international efforts to cut down emissions of PCBs have been effective to reduce concentrations in open water ecosystems like the BCZ but had little effect in the urbanized and industrialized area of the Scheldt estuary. Most likely, estuaries are subject to secondary emissions from historical pollution. In chapter III, trends found for the BCZ (chapter II) are confirmed at larger spatiotemporal scales. In chapter III multidecadal field observations (1979–2012) in the North Sea and Celtic Sea are analyzed to infer spatiotemporal concentration trends of PCBs in mussels (Mytilus edulis) and in sediments. Decreasing interannual PCB concentrations are found in North Sea sediments and mussels. PCB concentrations in sediments show, less than PCB levels in mussels, decreasing interannual trends. In addition in chapter III, interannual changes of PCB concentrations are separated from seasonal variability. By doing so, superimposed to the generally decreasing interannual trends, seasonally variable PCB concentrations are observed. These seasonal variations are tightly coupled with seasonally variable chlorophyll a concentrations and organic carbon concentrations. Indeed, the timing of phytoplankton blooms in spring and autumn corresponds to the annual maxima of the organic carbon content and the PCB concentrations in sediments. These results demonstrate the role of seasonal phytoplankton dynamics (biological pump) in the environmental fate of PCBs at large spatiotemporal scales.The latter is a novel result since the working of the biological pump was never assessed before based on field data collected at the scale of a regional sea in multiple decades.Despite the generally decreasing spatiotemporal trends of PCBs that are found in chapter II and III, it is not clear whether current concentrations (still) pose a risk to marine ecosystems. In chapter IV, the spatiotemporal trends inferred in chapter III are used to assess the ecological risk of PCBs in North Sea and Celtic Sea sediments and mussels. To do so, PCB concentrations are compared with environmental assessment criteria (EAC). It is found that the potential ecotoxicological risk of PCBs change considerably over time and in space. Risk quotients (RQs) of PCBs in marine sediments primarily depend on the location of the monitoring site, i.e. the closer to the coast, the higher the RQ. Especially in summer, when PCB concentrations in sediments are high, PCBs present in marine coastal sediments may pose an environmental risk. By contrast, RQs in mussel depend first on the interannual changes of PCB concentrations. At present, in the Celtic Sea, RQs in mussels are below the value of 1, suggesting no potential environmental risk. In the North Sea, however, PCBS in mussels may still exceed the prescribed environmental quality criteria. Overall, the results shown in chapter IV demonstrate that the spatiotemporal variability in PCB concentrations should be considered in future environmental risk assessments.Comparing concentrations of chemicals with quality thresholds (as in chapter IV) only suggest a potential ecological risk. Therefore, in case if risk quotients exceed the value of 1, additional assessments are recommended. Considering the results obtained in chapter IV, in chapter V, additional experimental studies are performed in which a marine diatom is exposed to a realistic mixture of organic contaminants. To do so, passive samplers are used to achieve exposure to realistic mixtures of organic chemicals close to ambient concentrations. The main conclusion is that organic chemicals present in Belgian marine waters do not affect the intrinsic growth rate of Phaeodactylum tricornutum. In this context, caution is needed when extrapolating these results to field conditions. In the present research, results were obtained under laboratory controlled conditions with one single species and thus neglecting possible species interactions. Therefore, prior to extrapolating these results to other diatoms and other groups of phytoplankton species, it is suggested to assess the validity of the results in a mesocosm experiment (including multiple species and different trophic levels) or under field conditions. In addition, in chapter V, the relative contribution of organic chemicals to the growth of a marine diatom is examined. Natural drivers such as nutrients regime, light intensity and temperature explain about 85% of the observed variability in the experimental data.Although the methodology used in chapter V is a standard way to assess toxicity of chemicals, it is not realistic to use just one algal species to represent ecotoxicological effects of an entire phytoplankton community. Therefore in chapter VI, an ecosystem model is used to assess the potential adverse effects of organic contaminants on the total primary production. To do so, we model phytoplankton dynamics using four classical drivers (light and nutrient availability, temperature and zooplankton grazing) and test whether extending this model with a POP-induced phytoplankton growth limitation term improves model fit. As inclusion of monitored concentrations of PCBs and pesticides did not lead to a better model fit, it is suggested that POP-induced growth limitation of marine phytoplankton in the North Sea and the Kattegat is small compared to the limitations caused by the classical drivers. The inferred contribution of POPs to phytoplankton growth limitation is about 1% in Belgian coastal waters, but in the Kattegat POPs explain about 10% of the phytoplankton growth limitation. These results suggest that there are regional differences in the contribution of POPs to the phytoplankton growth limitation. The validity of these conclusions should be further assessed for other substances, other species and higher trophic levels.