The first studies of diatoms and river pollution were carried out 60 years ago, and the suitability of these micro-algae as bioassessment indicators for monitoring river quality was quickly demonstrated. The objective of this review is to describe the state of the art of the study of diatoms and river pollution between 1999 and 2009. The objective was to group the publications homogeneously according to subject and to indicate which subjects never overlap. To this end a lexical analysis was conducted on the abstract structure of these publications and seven K-means clusters were defined. Most of the items in this area were found to have been published in Hydrobiologia. One group (group 6) comprises publications about a discipline (ecotoxicology); several others group publications on the basis of the same approaches, for example species (group 3), basin (group 7), or spatial approaches (group 2). Other publications are brought together because the studies used a common method of using data, namely predictive models (group 1) or biotic indices (group 4). One group of publications was of studies performed in the same area—South Africa (group 5). Several remarks can be made. First, ecotoxicological studies are mostly experimental and restricted to small study areas. To answer society’s demand for new assessment tools for micropollutant assessment, the next step would be to have more in-situ tests on larger spatial scales. Second, diatom biomonitoring uses the word “species” extensively, because this is the basis for establishing the lists of flora which are used extensively in such subject areas. Species is closely related to taxonomy; nevertheless this discipline is very rarely addressed in the papers. Third, phylogeny is never addressed in the publications. This is significant because phylogenetic studies for freshwater macroinvertebrates enable appropriate definition of taxonomic aggregations that can be used as accurate indicators of particular environmental stressors.