The dynamics of foot-and-mouth disease epidemics in France was examined through simulations based on an extended state-transition model. Contagion modelling depended on specific parameters: the so-called dissemination rates. Estimation of these parameters relied on a specific discrete-event simulation model. We were able to address the problem of the hidden spread of the epidemic, before the first outbreak is diagnosed. Furthermore, we took into account the silent development of the disease in affected herds before the diagnosis. The effect of control measures such as the active search for secondary outbreaks could thus be studied. We used the model to compare the development of FMD epidemics in two very different French regions and for different control strategies implemented by the animal-health authorities. These strategies gave similar results in a low herd density area, whereas in a high herd density area, the slaughter of contact herds greatly improved the stamping-out strategy. Finally, key parameters of the model were detected through a sensitivity analysis.