Ecologists and epidemiologists worry that global warming will increase disease prevalence. These fears arise because several direct and indirect mechanisms link warming to disease, and because parasite outbreaks are increasing in many taxa. However, this outcome is not a foregone conclusion, as physiological and community-interaction-based mechanisms may inhibit epidemics at warmer temperatures. Here, we explore this thermal-community-ecology-based mechanism, centering on fish predators that selectively prey upon Daphnia infected with a fungal parasite. We used an interplay between a simple model built around this system's biology and laboratory experiments designed to parameterize the model. Through this data-model interaction, we found that a given density of predators can inhibit epidemics as temperatures rise when thermal physiology of the predator scales more steeply than that of the host. This case is met in our fish-Daphnia-fungus system. Furthermore, the combination of steeply scaling parasite physiology and predation-induced mortality can inhibit epidemics at lower temperatures. This effect may terminate fungal epidemics of Daphnia as lakes cool in autumn. Thus, predation and physiology could constrain epidemics to intermediate temperatures (a pattern that we see in our system). More generally, these results accentuate the possibility that warmer temperatures might actually enhance predator control of parasites.