As was shown by a series of studies carried out for the most part in marine environments in the eighties, it is possible to describe phytoplankton composition using carotenoid pigment signatures specific of the different algal classes, and analyzed by HPLC. The most popular calculation method is based on multiple regression algorithms, and yields a series of ratios between chlorophyll a and the various carotenoids of interest. The final result of such calculations is an estimate of the biomass of all algal classes, expressed in chlorophyll a 'equivalents'. Coupled with other methods, analyzing pigments makes it possible to setup nearly automatic analytic procedures capable of handling large sample sets, and therefore to carry out large-scale census studies or else finely detailed studies of spatio-temporal variations in composition of algal assemblages. Quantifying pigments also renders the detection of phototrophic bacteria feasible, in deep anoxic waters. The validity of this technique can be proven, most notably by the similitudes between chl a/xanthophyll ratios obtained from marine, estuary, and more seldom lacustrine environments. This method based on major biomarkers however has its limitations, as it cannot discriminate between algal classes with comparable pigment signatures, like diatoms and Chrysophyta for exemple. An other approach can help resolve that problem, namely the CHEMTAX software package. Designed by australian oceanographers, it offers the possibility to estimate biomass on the basis of several pigments at once for each algal class. Its performances have been tested on data sets from various plankton collections from lakes of the LTER (North Wisconsin, USA, Descy et al 2000), the River Meuse in Belgium (1997 and 1998), the reservoirs of Bütgenbach, Robertville and Esch-sur-Sûre.