Following the publication of the Approved Lists, there has been a tendency to regard all subsequent revisions of classification as providing improved nomenclature, to be accepted without question. This takes no account of the fact that such revisions may be based on one of three alternative concepts, phenetic, phylogenetic or polyphasic classification, sometimes leading to different, valid, but incompatible nomenclature, or that some investigations are based only on subsets of relevant taxa and on limited data, leading to incomplete and sometimes confusing revisions of nomenclature. The polyphasic approach to classification has widespread support, although there appears to be a tendency to allow comparative sequence analyses of 16S rDNA to determine classification contrary to the indications of other data. In some cases, classification is based solely on 16S rDNA data. Examples are considered. Consideration is given to the criteria by which taxa are circumscribed, particularly at the level of genus and species. It is suggested that there is a need for reconciliation of the criteria by which taxa at these levels are circumscribed. Recent studies demonstrating the widespread occurrence of horizontal gene transfer suggest that there is a need for caution in monophyletic interpretations, especially when these are based on the analysis of single sequences.