Carrier proteins are important for improving the efficiency of synthetic peptide vaccines. Recently, genetically-based systems such as filamentous phage display or glutathione S-transferase (GST)-fusion proteins have been employed for immunisation. Whilst these carrier systems can facilitate the evaluation of a potential vaccine by reducing the time and cost of production, their relative efficacy and the kinetics of the immune response to each carrier has not been directly compared. In this study, we have displayed the epitopes of the anti-ErbB-2 Mabs N12 (C531-A586, EP531) and N28 (T216-C235, EP216) on phage minor coat protein pIII, major coat protein pVIII and GST. Balb/c H-2(d) mice were immunised with the constructs and the sera were tested after the initial, the 3rd, 5th and 6th immunisations for an anti-peptide, an anti-ErbB-2 and an anti-carrier response. The specificity of the antibody response was also mapped using synthetic peptides. It was found that GST was the best of the three carriers, both in terms of the magnitude and the kinetics of the induced anti-peptide and anti-ErbB-2 response. Multiple (five) administrations of the immunogens were necessary to obtain a high titre of antibodies specific for ErbB-2. It was further noted that whilst an anti-EP531 response was induced using all three carriers, EP216 was not immunogenic irrespective of the carrier used. The lack of immunogenicity of EP216 implies it does not contain a H-2(d) T cell recognition site. All three carriers provide a useful system for vaccination and consequently facilitate the identification of T-cell epitopes in Balb/c inbred mice.