The genetic variation in antibody responses of mice to glutathione S-transferase (GST) enzymes of Schistosoma japonicum worms, and in particular to a Mr 26,000 species termed Sj26, was analysed. Sera from infected mice, or mice immunized with adjuvant and an Sj26 β-galactosidase fusion protein produced in Escherichia coli (Sj26FP), or purified near-native recombinant Sj26 produced in E. coli (rSj26), were assayed by enzyme-linked immunosorbent assay (ELISA) for antibody titres to GST purified from adult worms. Anti-GST antibody levels are high in a mouse strain, WEHI 129/J, that is genetically resistant to infection with S.japonicum. Antibody responses to GST are low in BALB/c mice and intermediate in most other mouse strains analysed such as CBA/H and C57Bl/6. Responsiveness to Sj26 in adjuvant is dominant in (BALB/c x WEHI 129/J)F1 hybrids. BALB/c.H-2b and BALB/c.H-2(k) mice are higher responders than BALB/c. One feature of antibody responses to Sj26 is the variability within a group of mice. When rSj26 conjugated to the hapten azobenzenearsonate was used as immunogen, BALB/c mice produced substantial amounts of anti-Sj26 antibodies. In an attempt to correlate antibody levels with resistance in infected mice, a new functional assay was devised to measure the ability of sera to inhibit the binding of rSj26 to glutathione. However, there was no correlation between inhibitory titre in this assay and the numbers of worms recovered. In regard to the candidacy of GST as a vaccinating antigen in schistosomiasis japonica, the data raise the problem of variable responsiveness to the antigen that will need to be overcome by antigen modification and/or powerful adjuvants.