We used long-term data collected for up to 10 yr (1994–2004) at 23 trapping arrays (i.e., webs and grids) in Arizona, Colorado, Montana, and New Mexico to examine demographic factors known or suspected to be associated with risk of infection with Sin Nombre virus (SNV) in its natural host, the deer mouse (Peromyscus maniculatus). Gender, age (mass), wounds or scars, season, and local relative population densities were statistically associated with the period prevalence of antibody (used as a marker of infection) to SNV in host populations. Nevertheless, antibody prevalence and some of the risk factors associated with antibody prevalence, such as relative population density, gender bias, and prevalence of wounding, varied significantly among sites and even between nearby trapping arrays at a single site. This suggests that local microsite-specific differences play an important role in determining relative risk of infection by SNV in rodents and, consequently, in humans. Deer mouse relative population density varied among sites and was positively and statistically associated with infection prevalence, an association that researchers conducting shorter-term studies failed to demonstrate. Both wounding and antibody prevalence increased with mass class in both males and females; this increase was much more pronounced in males than in females and wounding was more frequent in adult males than in adult females. Prevalence of wounding was greatest among seropositive deer mice, regardless of mass class, but many deer mice without detectable wounds or scars eventually became infected. Many of these patterns, which will be useful in the development of predictive models of disease risk to humans, were only detected through the application of data collected over a long (10-yr) period and with abundant replication.