Fish, like other vertebrates, posses an e$cient immune system based on humoral and cellular reactions (Manning, 1994). An antibody response is mounted in fish after immunisation with acellular antigens (Joosten et al., 1997; Palm et al., 1998) and in response to vaccination (Boesen et al., 1997; Marsden & Secombes, 1997; Palm et al., 1998). The kinetics of production of antigen-specific immunoglobulins (Ig) in teleosts shows a primary and secondary humoral response (van Muiswinkel, 1995). The cells responsible for antibody production are plasma cells and B cells (Koumans-Van Diepen et al., 1994), and they have traditionally been enumerated by the plaque-forming assay (PFA) (Jerne & Nordin, 1963; Rijkers et al., 1980). In recent years, the evaluation of antibody-secreting cells has been conveniently achieved using the ELISPOT assay in fish (Boesen et al., 1997; Joosten et al., 1997; Bandin et al., 1997; Davidson et al., 1997), as well as in other vertebrates (Wu et al., 1998; Chen et al., 1995; Hogenesch et al., 1996; Yoo et al., 1998; Kim et al., 1999). This assay can also be used to study Ig synthesis in fish in response to environmental pollutants (Aaltonen et al., 1997), and in mammals has been adapted to detect individual cytokine-secreting cells (Vandermeide et al., 1995). The aim of the present work was to study the humoral immune response of the Mediterranean sea bass Dicentrarchus labrax by means of a simplified ELISPOT that would allow quantitative data on B cell activities in vitro (production of antigenspecific Ig) without manual counting of the read out at a microscope. Such an assay would allow manipulation of a large number of samples and reduce assay time. Sea bass were produced and reared in seawater aquaculture in a local fish farm (La rosa S.r.l., Orbetello, Italy). Eighteen month old animals, 298 76 g in weight, were used in all experiments. Prior to handling, fish were anaesthetised with 1 mg ml 1 of tricaine methanesulphonate (Sigma). Fish (n=6) were injected intraperitoneally (i.p.) with 250 l of 0·15 M NaCl, 2·7 mM KCI, 1·5 mM KH2PO4 and 8 mM Na2HPO4, pH 7·4 (PBS) containing 250 g of dinitrophenylated keyhole limpet haemocyanin (DNP-KLH, Sigma) emulsified with 100 l of Freund’s complete adjuvant (Serva, Heidelberg, Germany) (Findlay & Tatner, 1996). Alternatively, fish (n=4) were injected i.p. with 250 l fish 1 of PBS containing bacteria centrifuged from a 0·5 ml solution (ca. 2·5 10 cells) of Vibrio Anguillarum serotype 1 and serotype 2 vaccine (Microtek Europe, Brambles, U.K.) without adjuvant. Control fish (n=10) were injected i.p. with