The first large efficacy trial of a vaccine for the tropical disease dengue has returned puzzling results. The vaccine, developed by Sanofi Pasteur, proved safe, had an overall efficacy rate of 56.5%, and reduced cases of severe disease by 88.5%. But it had limited efficacy against dengue 2, one of four dengue virus serotypes in circulation. First infections are typically mild; but for poorly understood reasons, infection with a second serotype can lead to severe disease. Dengue is only occasionally fatal, but the World Health Organization estimates that severe dengue sends 500,000 people to hospitals each year, straining health care systems. In considering the vaccine for national immunization programs, public health officials will have to carefully weigh the benefits of reducing that burden against the yet-to-be-determined cost of the vaccine and its limited efficacy against dengue 2. The results also pose quandaries for scientists. The trial showed that a standard in vitro assay used to predict the level of immunity in vaccines correlated poorly with actual protection from disease. This poses a challenge for other groups working on dengue vaccines as they are all using this standard test to gauge efficacy.