Air travel has a decisive role in the spread of infectious diseases at the global level. We present a methodology applied during the early stages of the COVID-19 pandemic that uses detailed aviation data at the final destination level in order to measure the risk of the disease spreading outside China. The approach proved to be successful in terms of identifying countries with a high risk of infected travellers and as a tool to monitor the evolution of the pandemic in different countries. The high number of undetected or asymptomatic cases of COVID-19, however, limits the capacity of the approach to model the full dynamics. As a result, the risk for countries with a low number of passengers from Hubei province appeared as low. Globalization and international aviation connectivity allow travel times that are much shorter than the incubation period of infectious diseases, a fact that raises the question of how to react in a potential new pandemic.Int. J. Environ. Res. Public Health 2020, 17, 3356 2 of 15 information and air travel networks to produce a user-defined global map of risk distribution [9]. A generic tool that allows the estimation of the median early disease arrival time from around the world using air transport schedules was proposed by [10]. Air travel patterns were also employed by Bajardi et al. [11] in order to theoretically model the potential impacts of travel restrictions on the spread of an epidemic. Nevertheless, while travel and trade have been shown as relevant for the international spread of the 2009 A/H1N1 influenza, the slow deployment of control measures in countries with lower healthcare capacities led to spatial imbalances [12]. Air transport data have been used indirectly to measure the effective distance and the relative arrival time of a disease outbreak, with promising results when tested on historical data [13]. Based on the same premises, producing near real-time or now-casting predictions on the international spread of COVID-19 appeared to be based on scientifically sound principles [14].However, most expectations about the ability to predict the global spread of COVID-19 proved misleading or of limited use.This includes our own work, which we present here in order to discuss the reasons why most projections missed important aspects of the disease propagation dynamics. In our opinion, and as a part of the scientific process, it is important to publish and discuss even the negative results of research activities. The lessons learned from unsuccessful applications can be valuable for future work, either by improving current practices or by re-examining risk assessment from a new perspective. Especially in the case of predictive modeling, the wide body of literature that presents successful ex-post examples of possible value may give the false impression that the spatial aspects of disease propagation are more easily predictable than what is actually possible in reality.The question this paper is trying to answer is whether air passenger traffic alone can provide early information...