The possibility of identifying the bacteria in a sewage-sludge anaerobic digester by using the polymerase chain reaction to amplify 16S rRNA sequences from digester contents and comparing these with a general database of sequences was investigated. The database comprised bacteria from many ecosystems but not from digesters. No digester sequences were completely identical with those in the database. However, the database sequences most closely corresponding were representatives of genera, or in some cases species, of bacteria which had been previously cultured from digesters, or might have been present in the feedstock of the experimental digester. Digester floras are known to be heterogeneous and it is suggested that it is more likely that the sequences were from strains of digester bacteria related to, but not exactly the same as, bacteria in the database rather than being from entirely unknown bacteria. While exact identification would depend on comparisons with data from digester isolates, comparisons with a general database might be used to tentatively identify bacteria homologous to those in digesters and possibly to follow changes with time in a digester flora.