Mucosal and systemic antibody production in rainbow trout, Oncorhynchus mykiss (Walbaum), was evaluated following different antigen delivery routes. A BIAcore instrument (Pharmacia) allowed direct detection of antibody-antigen interactions by surface plasmon resonance changes. These interactions were measured in real-time without secondary reagents or extraneous labels. Groups of rainbow trout were immunised with a hapten-carrier antigen consisting of fluorescein isothiocyanate (FITC) conjugated to keyhole limpet haemocyanin (KLH) or phosphate buffered saline (PBS) pH 7.2. Antigens were administered intraperitoneally (i.p.) with or without Freund's complete adjuvant (FCA) or peranally (p.a.) directly to the gastrointestinal (GI) tract. Serum and mucosal anti-FITC responses were significantly (P<0.05) higher in FITC-KLH/FCA groups, clearly showing that adjuvant incorporation enhances mucosal as well as sytemic immunity. Antigen uptake and processing in fish immunised p.a. and i.p. without adjuvant was much less efficient and resulted in relatively low levels of serum and mucosal antibody production. Interestingly, mucosal responses in these groups peaked prior to serum responses suggesting possible early stimulation of mucosal defences. Mucosal antibody production in fish receiving FITC-KLH/FCA correlated more closely with serum responses, indicating possible transfer of serum derived antibody to mucosal sites. Mucosal and serum responses were confirmed as immunoglobulin (Ig) by subsequent reactivity with an anti-trout serum IgM monoclonal antibody (1.14) passed over flow cells containing anti-FITC antibodies. Further analysis showed significantly lower (P<0.05) reactivity of early mucus anti-FITC components (4 weeks post-immunisation) to 1.14. Purified serum and mucus Ig from non-immunised fish showed different protein banding patterns by SDS-PAGE under reducing conditions. Immunoblotting with 1.14 also showed weak reactivity to mucus Ig in control fish while reacting strongly to mucus Ig from immunised fish. These data suggest that early mucosal responses in trout may consist of heterogeneous forms of Ig differing in characteristics to serum Ig. BIAcore analysis in this context and as a means of measuring antibody response proved useful, and has the potential to become a valuable new tool in the study of fish immunology.