Biodiversity loss may increase the risk of infectious disease in a phenomenon known as the dilution effect. Circumstances that increase the likelihood of disease dilution are: (i) when hosts vary in their competence, and (ii) when communities disassemble predictably, such that the least competent hosts are the most likely to go extinct. Despite the central role of competence in diversity–disease theory, we lack a clear understanding of the factors underlying competence, as well as the drivers and extent of its variation. Our perspective piece encourages a mechanistic understanding of competence and a deeper consideration of its role in diversity–disease relationships. We outline current evidence, emerging questions and future directions regarding the basis of competence, its definition and measurement, the roots of its variation and its role in the community ecology of infectious disease.