A large number of groups at the Institut Pasteur perform state-of-the-art research documenting the dynamics of infection in vivo for a variety of pathogens. These dynamics are often complex and the data may not always allow to easily assess mechanisms at stake or estimate key biological parameters. The parallel with the difficulties epidemiologists are confronted to when studying epidemics dynamics at the level of a population is striking. We believe that, just like in epidemiology, the use of sound mathematical and statistical models of the infection process may substantially strengthen interpretation of these data. Given the huge potential for many groups within the Institute, we aim to invest substantial efforts in this research area. We have already started modelling the dynamics of Listeria infection with Olivier Disson and Marc Lecuit, with the view to better understand bacterial access to the central nervous system. We expect to start similar projects on chikungunya and dengue in a near future. The very diverse set of outstanding competences at Institut Pasteur offers a unique opportunity to develop a multidisciplinary perspective on infectious disease modelling and construct the next generation of mathematic models that make best use of data gathered at very different scales (cellular, host, population).