Using a complete series of overlapping peptides, we have identified the T cell epitopes of a malaria vaccine candidate, the circumsporozoite (CS) protein, that are recognized by sporozoite-exposed residents of a non-endemic country. This protein and subunits from it are being considered as malaria sporozoite vaccine candidates, as CS-specific antibodies and cytotoxic T lymphocytes have been shown to have a role in protection. The rationale for developing an antibody-based vaccine is that in Plasmodium falciparum the immunodominant B cell epitope of the protein, (Asn-Ala-Asn-Pro)n [(NANP)n], is invariant. However, the ideal vaccine must contain CS protein-derived T cell antigenic epitopes to allow natural boosting of the antibody response following sporozoite exposure. Here, we show that major differences occur between the CS-specific T cell responses of non-endemic Caucasians and an endemic African population. HLA differences between the populations are, in part, responsible. Subunit malaria vaccines for one population may be ineffective in a different population.