Enterobacter sakazakii is an emerging pathogen associated with life-threatening neonatal infections resulting from the consumption of contaminated powdered infant formula milk (IFM). Recent taxonomic analyses have determined that E. sakazakii comprises a number of genomospecies, and it has been proposed that E. sakazakii be reclassified as a novel genus, "Cronobacter". Accurate methods are required for the rapid detection and identification of this group of micro-organisms, since even low cell numbers have been reported to cause disease. The aim of this study was to evaluate various E. sakazakii detection methods in order to ascertain the most suitable method for detection and identification of these pathogenic agents. Samples from IFM and the environment were evaluated for the presence of E. sakazakii using the isolation steps (pre-enrichment, enrichment and selection) described in the Food and Drug Administration (FDA) method for E. sakazakii detection. Sixty-four isolates (50 from IFM and 14 from the environment) were selected from tryptone soy agar (TSA), regardless of colony appearance, and these isolates were identified by 16S ribosomal DNA (rDNA) sequencing. Thereafter, different culture-dependent and culture-independent methods were evaluated to accurately detect and identify the E. sakazakii isolates. These methods included the assessment of yellow pigment production on TSA, typical colonies on chromogenic Druggan-Forsythe-Iversen (DFI) and Chromocult Enterobacter sakazakii (CES) media and polymerase chain reaction (PCR) using six different species-specific primer pairs described in the literature. Identification of E. sakazakii using yellow pigment production was demonstrated to have a low sensitivity, specificity and accuracy (87%, 71% and 74%, respectively), which lowers the suitability of the FDA method. Chromogenic DFI and CES media were sensitive, specific and accurate (100%, 98% and 98%, respectively) for the detection of E. sakazakii. The specificity of the PCR amplifications ranged from 8% to 92%, emphasising the need for rigorous primer testing against closely related species. Of the primer pairs evaluated, Esakf/Esakr were the most suitable for E. sakazakii detection and identification. The detection limit of Esakf/Esakr was found to be 10(4) CFU/ml. This study demonstrated that no single method was capable of unambiguously confirming the presence and identity of E. sakazakii isolates, that each method had inherent advantages and disadvantages, and that in most cases several methods were required for accurate detection and identification. Further, it was demonstrated that the current FDA method for E. sakazakii detection should be revised in the light of the availability of more sensitive, specific and accurate detection methods.