A few dozen models have been developed for predicting the risk or intensity of Fusarium head blight (FHB) in wheat, mostly commonly by correlating weather variables with a disease variable. These statistical models have played an important role in disease management but are not suitable for understanding which and how different factors affect the behaviour of the epidemics. A few mechanistic FHB models, which link equations that describe epidemic processes (infection risk, inoculum risk, etc.), allow to estimate a daily risk which accumulates during the season. We developed a compartmental model, also known as SEIR (Susceptible, Exposed, Infectious, Removed), for simulating FHB epidemics. A series of ordinary differential equations are linked to estimate the amount of susceptible host sites (spikelets), based on degree-day accumulation, which move through the other three compartments. Two temperature-driven rates drive the amount of latently infected sites: rate of infection (anthers) and a rate of lesion expansion (spikelets within a wheat head). A time-decay model corrects for new infections by taking a spore cloud density into account. Model outputs resembled the typical FHB progress curves. The structure and expansions (host resistance and fungicide effects) of the model as well as a sensitivity analysis will be presented.