History shows that vaccines are most easily developed for those organisms that induce natural immunity after a single infection. For malaria, partial antiparasite immunity develops only after several years of endemic exposure. Evidence suggests that this inefficient induction of immunity is partly a result of antigenic polymorphism, poor immunogenicity of individual antigens, the ability of the parasite to interfere with the development of immune responses and to cause apoptosis of effector and memory T and B cells, and the interaction of maternal and neonatal immunity. Vaccine strategies that are likely to be ultimately successful are those that combine many antigens to induce a maximal response to protective determinants that might not be normally recognized following normal infection of naive individuals. Whole organismal approaches and the use of ultra-low doses of antigens have shown success in human and animal studies by inducing enhanced immune responses to multiple antigens. These, and related hypervalent subunit approaches, could lead to a viable vaccine.