A number of different schistosome antigens are capable of partially protecting experimental animals from challenge infection. More than 100 such antigens have been identified, about 15% of which are strongly protective and deemed promising though they do not reach the level close to sterile immunity seen after vaccination with irradiated cercariae. Studies of human correlate reactions, i.e. serological reactions and cytokine responses to schistosomiasis antigens, in individuals living in areas endemic for schistosomiasis have shown associations between certain antigen-specific immune responses and lack of re-infection over time. This approach was applied in Brazil and Egypt where it was possible to epidemiologically follow cohorts of individuals in endemic areas for extended periods of time correlating infection status with immune responses against a panel of well-researched, highly purified vaccine candidates. The immune correlates found were unique to each antigen and could be either positive or negative, i.e. associated with resistance or with susceptibility to re-infection. However, few antigens were clear-cut in this respect, i.e. the majority of them induced ambiguous responses. For example, a single antigen might have a significant positive correlation when antigen-driven interferon (INF)-gamma production is measured but also show a significant negative correlation with respect to the IgG1 titre induced. These observations suggest that there are desirable, antigen-specific immune responses that would be valuable in a vaccine but they also indicate that there are responses that must be avoided. The insights gained should be useful not only for antigen selection but also for vaccine formulation prior to Phase I/II trials in humans. It would be of great value if similar independent, long-term human correlate studies could also be undertaken in areas endemic for Schistosoma japonicum.