The dilution effect hypothesis posits that as species diversity decreases, infectious disease risk increases. Mechanisms behind diversity-disease relationships have been largely studied with vector borne and indirectly transmitted disease systems but less so with directly transmitted systems. The North American deermouse (Peromyscus maniculatus)-Sin Nombre hantavirus (SNV) system is an example of a directly transmitted disease system. In grasslands of western Montana, deermice compete with dominant voles (Microtus spp.) in species-poor small mammal communities. Empirical studies suggest that voles may influence SNV prevalence through effects on deermouse density, behavior or stress physiology. Therefore, our objective was to theoretically examine how vole density, instead of diversity per se, may impact SNV prevalence via competing effects on deermouse density, intraspecific contact rates and stress physiology. Using a previously validated epidemiological deermouse-SNV model, we explored various strengths of interspecific competition and stress under neutral, negative or positive relationships between deermouse contact rates and vole density. For a fixed range of vole densities, we found either a sole increase, a sole decrease, or a hump-shaped relationship (i.e. both increase and decrease) in SNV prevalence. A sole increase was observed only at lower competition strengths under neutral or positive relationships between contact rates and vole density. Our findings suggest that a hump-shaped relationship between competitor density and disease prevalence is theoretically possible across similar disease systems as a result of competing diversity-disease mechanisms.