I examined the effects of potential predators in relation to habitat selection in Peromyscus leucopus. I also examined habitat selection in three other sympatric species (Blarina brevicauda, Blarina carolinensis, Sorex longirostris). I utilized data fiom 49 sampling sites on National Park Service land; Petersburg National Battlefield Eastern Front Unit (15), Five-Forks Unit (17), Appomattox Court House National Historical Park (15) and Booker T. Washington National Monument (12). Sites were categorized by location (park unit) and habitat type (i.e. bottom-land hardwood), and the microhabitat within each sampling site was characterized by four variables (% cover of grasses, herbs, shrubs, and volume of downed wood). Importance values of tree species within each sampling site were also measured. Peromyscus leucopus, B. brevicauda, B. carolinensis, and S. longirostris were captured in all habitat types. Low capture rates for shrew species and high variability in the abundance of shrew species among all sampling sites were likely responsible for the lack of differences in abundance among habitats (p > 0.05). Peromyscus leucopus represented 76% of the captures among prey species and was the only species to demonstrate differences in relative abundance among habitat types. The relative abundance of P. leucopus was higher in bottom-land hardwood habitat when compared to pine forest plantation habitat (ANOVA, p P. leucopus among these habitat types may be explained by differences in understory structural diversity. Furthermore, the corresponding increase in the relative abundance of Procyon lotor (the most abundant predator during the study) in structurally heterogeneous habitat (bottom-land hardwood) suggests the importance of anti-predator behaviors within these habitat types. Ultimately, this may suggest that prey cannot escape the presence or calculate the abundance of predators and instead simply avoid dangerous habitats. Although the relative abundance of the most abundant predator (P. lotor) and prey (P. leucopus) species were positively associated within certain habitat types, a negative association between predator and prey species abundance was evident within parks. In the Eastern Front unit the relative abundance of prey (P. leucopus, S. longirostris) was lower in comparison to the Five-Forks unit while the abundance of P. lotor was higher (ANOVA, p P. lotor, Didelphis virginiana, and Urocyon cinereoargenteus were higher in the Eastern Front unit when compared to the Five-Forks unit. It appeared that the lethal effects of predators are evident at larger scales (within parks) despite the anti-predator behaviors of prey at smaller scales (within microhabitats). My results indicate that at larger scales (within parks) the lethal effects (removal of prey) of abundant predators may overwhelm the non-lethal effects (anti-predator behavior) exhibited by prey at smaller scales, especially in areas where structurally heterogenous habitats are lacking.