Studies in mice have shown that immunity to malaria sporozoites is mediated primarily by cytotoxic T lymphocytes (CTL) specific for epitopes within the circumsporozoite (CS) protein. Humans, however, had never been shown to generate CTL against any malaria or other parasite protein. The design of a sub-unit vaccine for humans relies on the epitopes recognized by CTL being identified and polymorphisms therein being defined. We have developed a novel technique using an entire series of overlapping synthetic peptides to define the epitopes of the Plasmodium falciparum CS protein recognized by human CTL and have analyzed the sequence variation of the protein with respect to the identified CTL epitopic domain. We have demonstrated that some humans can indeed generate CTL against the P. falciparum CS protein. Furthermore, the extent of variation observed for the CTL recognition domain is finite and the combination of peptides necessary for inclusion in a polyvalent vaccine may be small. If ways can be found to increase immune responsiveness, then a vaccine designed to stimulate CS protein-specific CTL activity may prevent malaria.