The ubiquity of biofilm and its classification as a microbial aggregate is discussed. Investigations into any microbial ecological problem operate at four levels: (i) in situ investigations, (ii) the use of microcosms, (iii) experimental model systems, and (iv) mathematical models. Each of these is defined and their use in biofilm research illustrated. It is concluded that all these approaches are valid and that scientific research in general and biofilm research in particular must profit by the use widely different methods if a complete understanding of a system is to be achieved.