Cyclic outbreaks of forest insects devastate forests, leading to widespread defoliation and tree death. Outbreaks would be far worse if not for epidemics of fatal virus diseases, which decimate outbreaking insect populations. The selection pressure imposed by these diseases suggests that natural selection may affect outbreaks, but understanding such effects is impossible with data alone. My lab has therefore used a combination of field experiments and models to test for effects of selection on outbreaks. Our work shows that both heritable host resistance and variation in viral virulence strongly affect outbreaks of the the gypsy moth, Lymantria dispar, an introduced pest of eastern hardwood forests in North America. Over the last few decades, however, an introduced fungal pathogen has competitively displaced the virus. The fungus provides better control, but its survival is much higher when the weather is cool and wet, whereas climate change is likely to cause weather conditions in the range of the gypsy moth to become increasingly hot and dry. By again combining models and data, we have shown that climate change will have a strong negative effect on the gypsy moth fungus, which may lead to the devastation of hardwood forests in North America. A key question is therefore, can the virus make a comeback? Our answers to this question are as yet incomplete, but provide initial chapters in an interesting story about the ecological effects of climate change.