Scrub typhus caused by the obligately intracellular bacterium, Orientia tsutsugamushi, is a major cause of life-threatening acute undifferentiated febrile illness in eastern Asia and the islands of the Western Pacific and Indian oceans. Since the estimation of an incidence of 1 million cases annually two decades ago, the number of cases has increased substantially in endemic regions, reappeared where the disease was forgotten, and spread northward. Trombiculid mites are both reservoir and vector. Despite 80 years of efforts to develop a vaccine, there is none. Protective immunity is mediated by antibodies and CD8 and CD4 T cells. Previous efforts have failed because of gaps in understanding immunity to O. tsutsugamushi, particularly the requirements for vaccine-induced immunity, lack of knowledge regarding immune memory in scrub typhus, and lack of attention to addressing the issue of cross-protection between strains. There are numerous strains of O. tsutsugamushi, and modestly durable immunity is strain-specific. Antibodies to the strain that caused infection are protective against challenges with the homologous strain but, despite reactivity with other immunodominant antigens, the immune serum does not protect against heterologous strains. Among the antigens detected by western immunoblot in immune sera (22-, 47-, 56-, 58-, and 110 kDa proteins), only the 56 kDa protein stimulates strong protection. This protein contains four hypervariable regions which are likely, on the basis of limited data, to be the targets of neutralizing antibodies. However, a method that definitively detects neutralizing antibody has yet to be developed. Only one study has used genomic data to pursue the discovery of protective antigens. Three conserved autotransporters were identified, and only immunization with ScaA provided protection against the homologous strain, but only 40% of animals were protected against challenge with a heterologous strain. A multiplex vaccine containing conformational antigens of the hypervariable regions of the 56 kDa protein of the strains of the greatest clinical and epidemiological importance, as well as conserved regions of the 56 kDa protein, ScaA, and other protective antigens identified by future genomic and bioinformatics methods should be developed and tested.