Nine similarity indices based on phytoplankton community structure were examined for their sensitivity to assess different levels of eutrophication. Two phytoplankton data sets, one from an open coastal system and one from a semi-enclosed gulf, associated with different nutrient dynamics and circulation patterns were used for evaluating the indices. The results have shown that similarity indices, measuring interspecific association and resemblance of phytoplankton communities between enriched areas and control sites, were effective for detecting spatial and temporal dissimilarities in coastal marine ecosystems. The structure of the oligotrophic habitat as a potential source of ambiguity for the results was discussed, whereas the validity ranges and the potential applicability of this method were deemed to be dependent on the size of the fraction of the common species among the samples, and the similarity of the classification patterns resulted from this subcategory and those extracted from the overall community data. Furthermore, the study provides a new technique based on the use of the "Box and Whisker Plot" designed to distinguish opportunistic and rare phytoplanktonic species. The similarity indices, applied solely to the dominant species abundance, were more sensitive to resolve eutrophic, mesotrophic and oligotrophic conditions. This procedure can be proposed as an effective methodology for water characterization and can also be used as a qualitative tracer for detecting renewal processes of coastal marine ecosystems.