Uncertainty was a defining feature of the Brazilian Zika crisis of 2015–2016. The cluster of cases of neonatal microcephaly detected in the country's northeast in the second half of 2015, and the possibility that a new virus transmitted by Aedes mosquitoes was responsible for this new syndrome, created a deep sense of shock and confusion in Brazil and around the world. When in February 2016 the WHO declared a Public Health Emergency of International Concern (PHEIC), it noted that it did so on the basis of what was not known about the virus and its pathogenic potential. To better understand the role that non-knowledge played in the unfolding of the Brazilian Zika crisis we differentiate between three different kinds of uncertainty: global health uncertainty, public health uncertainty, and clinical uncertainty. While these three forms of uncertainty were difficult to disentangle in the early weeks of the crisis, very soon each one began to trace a distinct trajectory. Global health uncertainty centered on the question of the causative link between Zika virus infection and congenital malformations, and was declared resolved by the time the PHEIC was lifted in November 2016. Public health and clinical uncertainty, in contrast, persisted over a longer period of time and did, in some important ways, become entrenched. This taxonomy of uncertainties allows us to explore the systematic nonproduction of knowledge in times of medical emergency, and suggests structural limitations in the framework of “emergency research” that global health institutions have developed to deal with unexpected threats.