This paper aims at discussing the two main modeling schemes that are used to describe dynamically the growth of microbial ecosystems, that are resource and density-dependent growth functions, respectively. Monod has been the first to hypothesize that this growth is, before all, an increasing saturated function of the main limiting substrate concentration. Contois assumed that the growth is not only a function of the substrate but also of the biomass density-itself, and thus the name « density-dependent ». In re examining their respective experiments (species used, conditions of experiments, mode of reactor functioning, measurement techniques), we try to understand the engines for a density-dependent phenomenon to appear. In particular, we refer to recent experiments where it was shown that density-dependent appeared as soon as the biomass structures into flocs or in the presence of filamentous bacteria even at relatively low concentrations. Based on this historical review of data, it is shown that density-dependent kinetics is not systematically a question of biomass density but rather related to its structure within the medium and to the mobility of microbial cells.