With the development of kinetic models and new design criteria for activated sludge systems, biomass determination becomes important. Various methods can be used to estimate the amount of biomass. Most of these methods are only applicable to pure culture studies. The activated sludge of wastewater treatment plants contains not only biomass, but also a high content of dead organic and inorganic material of unknown composition. Distinguishing between these sludge constituents is difficult. The best and most reproducible method for biomass estimation is often described as the determination of DNA content. This method includes the acid extraction of DNA, the quantitative determination of deoxyribose sugar by a colour reaction with diphenylamine, the calibration with standard DNA and the mathematical conversion into biomass. This study shows that the conventional method is strongly affected by unknown activated sludge constituents and in particular iron. The interference can be overcome by EDTA treatment. Inconsistencies in published calibration data are overcome.