Unidentified bacteria or isolates with ambiguous profiles.One of the most attractive potential uses of 16S rRNA gene sequence informatics is to provide genus and species identification for isolates that do not fit any recognized biochemical profiles, for strains generating only a “low likelihood” or “acceptable” identification according to commercial systems, or for taxa that are rarely associated with human infectious diseases. The cumulative results from a limited number of studies to date suggest that 16S rRNA gene sequencing provides genus identification in most cases (>90%) but less so with regard to species (65 to 83%), with from 1 to 14% of the isolates remaining unidentified after testing (5, 11, 17). Difficulties encountered in obtaining a genus and species identification include the recognition of novel taxa, too few sequences deposited in nucleotide databases, species sharing similar and/or identical 16S rRNA sequences, or nomenclature problems arising from multiple genomovars assigned to single species or complexes.Routine isolates.Surveys have looked at the feasibility of identifying routine clinical isolates or specific groups of medically important bacteria using SSU gene sequence data. In each of these studies, SSU sequence data has been compared to identification results obtained either in conventional or commercial test formats (Table 1). A couple of general observations can be made from these investigations, namely, (i) a higher percentage of species identifications were obtained using SSU sequence results than with either conventional or commercial methods and (ii) most studies, with the exception of one study by Fontana et al. (6), have found that 16S yielded species identification rates of 62 to 91%. In the study by Fontana et al. (6) the closest match in the MicroSeq 500 database was considered the identification no matter what the distance score was. For bacteria that are difficult to grow or identify the identification rates were lower with 16S rRNA sequencing (62 to 83%) than the values traditionally acceptable in the clinical laboratory (i.e., ≥90%) (12). Problems again revolved around complete and accurate databases and groups that are not easily distinguishable by 16S rRNA gene sequencing (2, 8).