An enzyme-linked immunosorbent assay (ELISA) to detect antibodies to Escherichia coli O157 lipopolysaccharide (LPS) was developed with sera from 63 children with confirmed recent E. coli O157 infection and from 256 age-stratified urban controls. The median ELISA values for control and case sera were 0.05 (interquartile range, 0 to 0.20; mean +/- standard deviation [SD], 0.15 +/- 0.22) and 1.41 (interquartile range, 1.11 to 1.59; mean +/- SD, 1.41 +/- 0.53), respectively (P < 0.001). With a breakpoint of 0.59 (mean ELISA value of the control sera + 2 SDs), the assay had a sensitivity, specificity, and positive and negative predictive values of 95, 94, 80, and 98%, respectively, for recent E. coli O157 infection. The O157 LPS assay and Vero cytotoxin (VT) 1-neutralizing-antibody (NAb) assay were used to compare the relative frequencies of O157 LPS antibodies and VT1-NAbs in an age-stratified urban population from Toronto, Ontario, Canada, and in 216 healthy family members from dairy farm in southern Ontario. The frequency of O157 LPS antibodies was about threefold higher in dairy farm residents (12.5%) than in urban residents (4.7%) (P < 0.01). Similarly, the frequency of VT1-NAbs was about sixfold higher in dairy farm residents (42.0%) than in urban residents (7.7%) (P < 0.001). These findings are consistent with a greater level of exposure of dairy farm residents to VT-producing E. coli (VTEC) strains. The high rate of seropositivity to VT1 in farm residents probably reflects the booster effect of repeated VTEC exposures and argues against a sustained generalized immunosuppressive effect of VT1. Seroepidemiological studies may help in assessing the level of exposure of different populations to VTEC strains.