The implementation of the European Water Framework Directive (WFD) requires the development of ecologically-based classification systems for anthropogenically-induced eutrophication in all types of water bodies. Due to the inherent high temporal and spatial variability of hydrological and geochemical parameters of the coastal waters of the southern Baltic Sea, discrimination between anthropogenic impact and natural variability is necessary. The development of statistical methods for this discrimination was the main aim of this study. These methods were used to derive indicative phytoplankton parameters for different stages of eutrophication for the investigation area. For this purpose, a long-term phytoplankton data series was analysed, which covered a broad salinity and eutrophication gradient. In order to detect eutrophication effects, the analysis was restricted to phytoplankton spring bloom events and to the salinity range between 5 and 10 psu, i.e. superimposing seasonal and hydrodynamic effects were eliminated. An artificial abiotic degradation vector was developed based on four typical water quality parameters. A total of 11 potentially indicative phytoplankton parameters on different taxonomical levels arose from a correlation analysis with this degradation vector. These indicators were then tested for their ability to discriminate between three eutrophication levels. Finally, seven phytoplankton indices could be proposed: total phytoplankton biovolume, the percentage of diatoms and the biovolume of different size ranges of diatoms and one indicative species (Woronichinia compacta).