There is growing recognition that the immune protection generated by prior pathogen exposure is often highly variable. Nonetheless, the role of prior exposure to pathogens in inducing population-level variation in host susceptibility, and how selection from pathogens alters such heterogeneity over time, remains largely unexplored. Further, the way in which host variation in susceptibility emerges from interactions between pathogens traits, such as virulence, and host traits, such as acquired protection, is key for predicting how pathogen strains will evolve and coexist in the breadth of host-pathogen systems characterized by incomplete protection. While theoretical models suggest key downstream consequences of host heterogeneity in susceptibility for pathogen dynamics and evolution, there have been few opportunities to empirically parameterize such models and test their predictions. Here we blend theory and experiments in a tractable natural system -- house finches and a common conjunctival bacterial pathogen -- to explore an under-considered but broadly relevant mechanism of heterogeneity in susceptibility: variable host protection generated from prior pathogen exposure. How such prior exposure induces or dampens population-level heterogeneity in susceptibility will depend on several factors, including 1) the extent of prior pathogen exposure that individuals experience during epidemics, and thus the variation in protection generated and then lost via waning immunity (ecological processes); and 2) the extent of pathogen-driven selection on host susceptibility in a given population over longer time scales (evolutionary processes). Heterogeneity in host susceptibility is also specific to a given pathogen, and thus can be influenced by traits of pathogens (e.g. virulence) that interact with host prior exposure. We will assess these theories-- and their epidemiological and evolutionary consequences-- among individual finches with varying histories of prior pathogen exposure, across populations with varying histories of pathogen-driven selection, and among pathogen strains that vary in virulence. The work iteratively merges empirical approaches with epidemiological and evolutionary models to test: 1) how prior exposure to pathogens, via both ecological and evolutionary processes, influences host heterogeneity in susceptibility, 2) the epidemiological consequences of exposure-induced heterogeneity, and, 3) how prior exposure interacts with pathogen traits to influence heterogeneity, strain coexistence, and virulence evolution. By doing so, the project will generate novel and diverse insights regarding a largely understudied but key form of host heterogeneity.