Graduation date: 2006I studied small-mammal communities and their response to grazing in mixed-conifer forests and oak woodlands in the Cascade-Siskiyou National Monument in southern Oregon. My objectives were to (1) compare small-mammal communities among forest types and grazing intensities, (2) identify riparian affiliated species, and (3) describe microhabitat associations. Over two years, I trapped 1,617 individual small mammals representing 20 species. Mixed-conifer forests had more species and greater abundances of small mammals than oak woodlands, but there were no significant differences between the forest types in species richness or diversity at the stand level, as most sites had 3 to 5 species at any given time. Evenness was higher in oak woodlands compared to mixed-conifer forests because of lower dominance by Peromyscus maniculatus and fewer rare species. P. maniculatus dominated both forest types. Neurotrichus gibbsii, Tamias amoenus, Tamias siskiyou, Neotoma cinerea, Clethrionomys californicus, Microtus californicus, Microtus longicaudus, and Microtus montanus were found only in mixed-conifer forests, whereas Peromyscus boylii were trapped exclusively in oak woodlands. Although the two forest types shared 10 species, Sorex trowbridgii and Peromyscus maniculatus were significantly (P < 0.05) more abundant in mixed-conifer forests, whereas Sorex vagrans, Reithrodontomys megalotis, Peromyscus truei, and Neotoma fuscipes were significantly more abundant in oak woodlands. Microtus townsendii was the only species that was equally abundant in both forest types, but was found exclusively within meadow sites in mixed-conifer forests. I found no significant differences in species richness, diversity, and evenness between heavy and lightly grazed sites. However, mean mammalian biomass was 138 g/ha lower in heavily grazed sites, and abundances were 89.3% of those in lightly grazed sites. Cumulative biomass and abundance across heavily grazed sites were 65% and 80%, respectively, of that in lightly grazed sites. Several relatively large species (e.g. Neotoma spp. and Microtus spp.) were less abundant or absent in heavily grazed sites. My results support previous studies that found lower abundances of small mammals associated with herbaceous cover (e.g. Microtus spp.) in heavily grazed sites and little or no effect for species associated with open habitats (e.g. Tamias spp.). I found significantly (P < 0.05) lower abundances of S. trowbridgii (conifer), Peromyscus spp. (oak), N. cinerea (conifer), N. fuscipes (oak), M. longicaudus (conifer), and M. californicus (conifer) in heavy versus lightly grazed sites. There were also fewer T. siskiyou, R. megalotis, and M. townsendii captured in heavy versus lightly grazed sites, but the differences were not significant (P > 0.05). S. vagrans and T. amoenus were significantly more abundant in heavily grazed sites. Sample sizes for the remaining species were too small for comparison (n < 6). N. gibbsii, C. californicus, and M. californicus were captured exclusively in lightly grazed sites; only T. amoenus was captured exclusively in a heavily grazed site. There was a significant (P < 0.05) decrease in the odds of catching R. megalotis, N. cinerea, N. fuscipes, and M. longicaudus on heavy versus lightly grazed sites after accounting for microhabitat associations and potential confounding factors. Odds of capture were significantly (P < 0.05) lower for P. maniculatus in heavy versus lightly grazed sites in oak woodlands, but there was less difference in abundance between grazing intensities in mixed-conifer forests. There was some evidence to suggest that grazing decreased the odds of capturing M. townsendii in riparian areas. Although I found significantly (P < 0.05) fewer S. trowbridgii in heavily grazed sites, my analysis indicated that microhabitat associations rather than grazing explained the difference. Grazing appeared to increase the use of riparian areas by Tamias spp. by removing herbaceous cover and increasing bare ground. I described microhabitat associations for 10 species and found a suite of species (S. trowbridgii, T. amoenus, T. siskiyou, N. cinerea) that were positively associated with coarse woody debris and/or snags. R. megalotis, M. longicaudus, and M. townsendii were affiliated with, but not obligated, to riparian areas. P. maniculatus were ubiquitous, but abundances were also higher in riparian areas compared with uplands. T. siskiyou and P. truei appeared to be associated with uplands rather than riparian areas. Although I captured few S. vagrans, N. gibbsii, M. californicus, and M. montanus, they were found in either riparian areas or around spring seeps, which support previous studies. My results supported the importance of riparian areas to small mammals as suggested by previous studies.