1. Plant defense against pathogens includes a range of mechanisms, including, but not limited to, genetic resistance, pathogen-antagonizing endophytes, and pathogen competitors. The relative importance of each mechanism can be expressed in a hierarchical view of defense. Several recent studies have shown that pathogen antagonism is inconsistently expressed within the plant defense hierarchy. Our hypothesis is that the hierarchy is governed by contingency rules that determine when and where antagonists reduce plant disease severity. 2. Here, we investigated whether pathogen competition influences pathogen antagonism using Populus as a model system. In three independent field experiments, we asked whether competition for leaf mesophyll cells between a Melampsora rust pathogen and a microscopic, eriophyid mite affects rust pathogen antagonism by fungal leaf endophytes. The rust pathogen has an annual, phenological disadvantage in competition with the mite because the rust pathogen must infect its secondary host in spring before infecting Populus. We varied mite-rust competition by utilizing Populus genotypes characterized by differential genetic resistance to the two organisms. We inoculated plants with endophtyes and allowed mites and rust to infect plants naturally. 3. Two contingency rules emerged from the three field experiments: 1) pathogen antagonism by endophytes can be preempted by host genes for resistance that suppress pathogen development, and 2) pathogen antagonism by endophtyes can secondarily be preempted by competitive exclusion of the rust by the mite. 4. Synthesis: Our results point to a Populus defense hierarchy with resistance genes on top, followed by pathogen competition, and finally pathogen antagonism by endophytes. We expect these rules will help to explain the variation in pathogen antagonism that is currently attributed to context dependency.