Aims: To develop methods to assess the efficiency of immunomagnetic separation (IMS). Methods and Results: The capturing efficiency of biosorbents for Escherichia coli O157:H7, constructed using streptavidin-coated magnetic beads and biotinylated antibodies, was tested using both in vivo and ATP bioluminescence. Both methods were suitable for the enumeration of bacteria captured by the biosorbents. The level of both ATP and in vivo bioluminescence depended on the media used, but was unaffected by the magnetic beads. The capture efficiency depended on time and sample volume, but did not depend on the length of spacer arm of the biotinylation agent. For cell concentrations of less than or equal to 10(5) cfu ml(-1) , in a 1-ml sample volume, nearly 80-85% recovery of the pathogen was observed after 0.5 h of incubation. For an 11-ml sample containing 10(4) cfu ml(-1) , maximum recovery (50% of cells) was achieved only after 2 h incubation. Conclusions: The detection limit of an ATP-based bioluminescent assay for E. coli O157:H7 was reduced by 1 log cycle after optimization of IMS. The bioluminescent methods could be used for screening and testing the affinity of antibodies or other affinity elements of biosorbents towards live bacterial cells. Significance and Impact of the Study: Bioluminescent assays provide an easy way to optimize conditions for the capture of bacteria by biosorbents in real time.