In recent years major efforts have been made to characterize parasite antigens thought to be suitable candidates for malaria vaccines. Many of the relevant plasmodial antigens have been found to contain extensive areas of short amino acid sequences organized in tandem repeats. These are usually strongly antigenic, forming linear epitopes seen by antibodies of the infected host. Several such epitopes have been identified and subunit vaccines are being designed in which synthetic peptides or gene constructs serve as immunogens. However, as an efficient malaria vaccine should give rise to anamnestic T-dependent antibody responses following reinfection after vaccination as well as to antibody independent cell-mediated immunity, efforts are now also being made to identify T-cell epitopes on the vaccine candidate antigens. In this paper the current Plasmodium falciparum sporozoite vaccines and the merozoite antigen Pf155/RESA, a possible candidate for a P. falciparum blood stage vaccine, serve as examples to illustrate recent advances made in this area as well as some of the problems remaining to be resolved.