Rapid, direct methods are needed to assess active bacterial populations in water and foods. Our objective was to determine the efficiency of bacterial detection by immunomagnetic separation (IMS) and the compatibility of IMS with cyanoditolyl tetrazolium chloride (CTC) incubation to determine respiratory activity, using the pathogen Escherichia coli O157:H7. Counterstaining with a specific fluorescein-conjugated anti-O157 antibody (FAb) following CTC incubation was used to allow confirmation and visualization of bacteria by epifluorescence microscopy. Broth-grown E. coli O157:H7 was used to inoculate fresh ground beef (<17% fat), sterile 0.1% peptone, or water. Inoculated meat was diluted and homogenized in a stomacher and then incubated with paramagnetic beads coated with anti-O157 specific antibody. After IMS, cells with magnetic beads attached were stained with CTC and then an anti-O157 antibody-fluorescein isothiocyanate conjugate and filtered for microscopic enumeration or solid-phase laser cytometry. Enumeration by laser scanning permitted detection of ca. 10 CFU/g of ground beef or <10 CFU/ml of liquid sample. With inoculated meat, the regression results for log-transformed respiring FAb-positive counts of cells recovered on beads versus sorbitol-negative plate counts in the inoculum were as follows: intercept = 1.06, slope = 0.89, and r2 = 0.95 (n = 13). The corresponding results for inoculated peptone were as follows: intercept = 0.67, slope = 0.88, and r2 = 0.98 (n = 24). Recovery of target bacteria on beads by the IMS-CTC-FAb method, compared with recovery by sorbitol MacConkey agar plating, yielded greater numbers (beef, 6.0 times; peptone, 3.0 times; water, 2.4 times). Thus, within 5 to 7 h, the IMS-CTC-FAb method detected greater numbers of E. coli O157 cells than were detected by plating. The results show that the IMS-CTC-FAb technique with enumeration by either fluorescence microscopy or solid-phase laser scanning cytometry gave results that compared favorably with plating following IMS.