In my last lecture I will focus on the issue of antiviral resistance during closed epidemics, again taking pandemic influenza as the paradigm. I will present new work which shows that previous assessments of the risk of antiviral resistance in influenza pandemics have been over-pessimistic, for 2 reasons: (a) previous simple models have often over-estimated the selection pressure imposed by antiviral treatment; (b) spread of new, rare phenotypes is dramatically slower when one accounts for host population structure than one would predict by assuming homogenous mixing. In addition, I will examine how the final impact of resistance during a closed epidemic depends on the transmissibility of a sensitive and resistant virus, the mutation rate from sensitive to resistant types and the level of seeding of both viral types at the start of the epidemic. Non-pharmaceutical public health interventions are shown to be able to either slow or hasten the spread of resistance depending on their effectiveness. Delaying use of antivirals or sequential use of different antivirals is shown to be effective at reducing the final impact of resistance. Overall, providing the strains seeding a pandemic in a country are predominantly drug-sensitive, I will argue that resistance is unlikely substantially reduce the effectiveness of antivirals during the first wave of a pandemic, but that intensive use of such drugs in the first wave can lead to a high frequency of resistance in later epidemics.