The specific humoral response of teleost fish to extracellular bacteria was examined using a rainbow trout -Vibrio anguillarurn model. Treatment groups were immunized by oral, immersion, and ~njection routes. All 3 delivery methods conferred full protection in controlled laboratory challenges (p < 0.01). Prior to boosting, serum antibody titers did not correlate with protection in the orally and irnmers~on-vaccinated groups, but, contrary to previous studies, titers measured 10 and 17 d after boosting correlated pos~tively with protection in all 3 vaccinated groups. The route of administration strongly affected the magnitude of the antibody response as measured by enzyme-linked immunosorbent assay (ELISA) and Western blots; however, the antigenic epitopes recognized were not substantially altered by delivery method as evidenced in lmmunoblot patterns. Given that the primary and booster vaccination protocols were identical, the data suggest that all 3 vaccinated groups may have had a specific humoral response following lnltial immunization but that specific serum antibody levels before boosting were toolow to be detected by ELISA In fish vaccinated by oral and immersion routes. An anamnestic response was evident in all 3 groups. The data support the possibility that teleosts, like higher vertebrates, have a protective immune response to extracellular bacteria that is predominantly humoral. Route of delivery may primarily affect the efficiency with which the immunogenic constituents of the vaccine are presented to the relevant recognition and effector components of the immune system.