We evaluated interspecific habitat use within a salt marsh small mammal assemblage on Mare Island, Solano County, California, USA, from 1989 to 1992, with emphasis on the endangered salt marsh harvest mice (Reithrodontomys raviventris). We livetrapped small mammals during 125 trap sessions at 20 different areas throughout Mare Island for a total of 55,189 trap-nights. We captured the salt marsh harvest mouse 4,147 times; the house mouse (Mus musculus), 1,936 times; the California vole (Microtus californicus), 372 times; and the shrew (Sorex spp.), 117 times, among 12,927 captures. We captured fewer than 10 rats (Rattus spp.), and we did not capture any western harvest mice (Reithrodontomys megalotis) or deer mice (Peromyscus maniculatus). We sampled vegetation characteristics at each trap location during 79 of 125 trap sessions for a total of 5,523 trap locations. During the summer, breeding, and fall seasons, habitats characterized by increased cover of forbs, particularly fat hen (Atriplex patula), were used to a greater extent by male than female salt marsh harvest mice. Both sexes of salt marsh harvest mice used areas with less cover of forbs, particularly fat hen and pickleweed (Salicornia virginica), during the winter, spring, postbreeding, and prebreeding seasons. House mice used habitats that were more patchily distributed or fragmented than salt marsh harvest mice. Habitat characteristics that were positively associated with salt marsh harvest mice tended to be negatively associated with house mice. Voles used habitats characterized by positive associations with shrub, pickleweed, litter, and woody debris cover, foliage height densities (0 to >40 cm), and mean vegetation height. Vole habitat use was negatively associated with water cover and depth. We suggest that reducing habitat patchiness throughout tidal marshes may reduce salt marsh harvest mouse competition with house mice, and restoring tidal action may reduce habitat competition with voles.