Despite efforts to develop vaccines that protect against malaria by inducing CD8+ T cells that kill infected hepatocytes, no subunit vaccine has been shown to circumvent the genetic restriction inherent in this approach, and little is known about the interaction of subunit vaccine-induced immune effectors and infected hepatocytes. We now report that immunization with plasmid DNA encoding the plasmodium yoelii circumsporozoite protein protected one of five strains of mice against malaria (H-2d, 75%); a PyHEP17 DNA vaccine protected three of the five strains (H-2a, 71%; H-2k, 54%; H-2d, 26%); and the combination protected 82% of H-2a, 90% of H-2k, and 88% of H-2d mice. Protection was absolutely dependent on CD8+ T cells, INF-gamma, or nitric oxide. These data introduce a new target of protective preerythrocytic immune responses, PyHEP 17 and its P. falciparum homologue, and provide a realistic perspective on the opportunities and challenges inherent in developing malaria vaccines that target the infected hepatocyte.