Although serum vibriocidal activity is used extensively as a marker of immunity to O1 Vibrio cholerae, there are limitations in this assay to detect instances of reexposure. We define the conditions operative in producing secondary vibriocidal responses in North American volunteers primed with either wild-type V. cholerae 1, 4, or 6 months later. Secondary serum vibriocidal responses occurred under two distinct secondary challenge conditions. The first occurred when secondary challenge produced a breakthrough in clinical protection. Following secondary exposure, 14 of 22 (64%) and 1 of 29 (3%) subjects with and without vibrio stool excretion, respectively, had secondary responses (P < 0.001); 5 of 6 (83%) and 10 of 45 (22%) subjects with or without diarrhea, respectively, mounted a secondary response (P = 0.006). The second condition occurred in the presence of full clinical protection but was dependent on the time interval between exposure. No subject (0 to 17) vaccinated with CVD 103-HgR and given homologous wild-type challenge within 4 months mounted a secondary vibriocidal response (P = 0.0009). The majority of the serum vibriocidal activity was of the immunoglobulin M (IgM) isotype, seen in 96 and 73% of subjects following primary and secondary exposure, respectively. Vibriocidal activity in the IgG fraction following primary and secondary exposures occurred with < or = 50% of volunteers; lipopolysaccharide (LPS)-specific IgG1 and IgG3 subclass responses supported the vibriocidal isotype data. However, following primary exposure, IgG4 LPS responses predominated, occurring in 81% of responding volunteers. These data suggest that, under certain conditions of secondary exposure to V. cholerae O1 antigens, when there is sufficient active local immunity present, there is a block of vibrio antigen resampling at the M cell level. We discuss the implications of and possible explanations for these findings.