The generalist fungal pathogen Pyrenophora semeniperda occurs primarily in cheatgrass (Bromus tectorum) seed banks, where it causes high mortality. We investigated the relationship between this pathogen and its cheatgrass host in the context of fire, asking whether burning would facilitate host escape from the pathogen or increase host vulnerability. We used a series of laboratory and field experiments to address the ability of host seeds and pathogen life stages to survive fire. First, we determined the thermal death point (TDP50; temperature causing 50% mortality) of seeds and pathogen propagules at two time intervals using a muffle furnace. We then measured peak fire temperatures in prescribed burns at sites in Utah and Washington and quantified seed and fungal propagule survival using pre- and postburn seed bank sampling and inoculum bioassays. Finally, we investigated the survival of both seeds and pathogen after wildfires. We found that radiant heat generated by both prescribed and wild cheatgrass monoculture fires was generally not sufficient to kill either host seeds or pathogen propagules; most mortality was apparently due to direct consumption by flames. The 5-min mean TDP50 was 164 C for pathogen propagules and 148 C for host seeds, indicating that the pathogen is more likely to survive fire than the seeds. Peak fire temperature at the surface in the prescribed burns averaged 130 C. Fire directly consumed 85-98% of the viable seed bank, but prescribed burns and wildfires generally did not lead to dramatic reductions in pathogen inoculum loads. We conclude that the net effect of fire on this pathosystem is not large. Rapid postburn recovery of both host and associated pathogen populations is the predicted outcome. Postfire management of residual cheatgrass seed banks should be facilitated by the persistent presence of this seed bank pathogen.