Bridging this data gap, and developing succinct yet realistic descriptions of epidemics’ impact on social-network dynamics, will prove key to making models of lethal epidemics more accurate. Three avenues of research need to be pursued. The first is controlled epidemiological intervention studies to determine how modifying contact networks affects disease transmission. Another will be the integrated collation and analysis of epidemiological and, preferably, quantitative social data from historical epidemics. Finally, protocols and data-collection systems should be designed to track the number of people infected per day in a future lethal outbreak as well as the behavioural response of the affected population.