Sampling biases resulting from type and placement of traps can strongly influence results of studies of small mammals. We compared trap success of modified Fitch and Sherman live-traps in riparian areas of the Oregon Coast Range. Of 8 species of small mammals captured in sufficient numbers to allow statistical comparisons, 6 (Peromyscus maniculatus, Tamias townsendii, Microtus orego?ai, Soreoc trowbridgii, S. pacificus/sonomae, and Neurotrichus gibbsii) were captured more frequently in Sherman than in Fitch traps. Microtus longicaudus was captured with equal frequency in the 2 trap types. Zapus trinotntus was captured most frequently in Fitch live-traps. Incidence of trap mortalities did not differ between the 2 types of traps. In general, Sherman live-traps are more effective than Fitch live-traps) but use of Fitch live-traps may be particularly useful in studies of Zapus trinota:tus. Studies of small mammals often rely on data collected using live-traps. As a result, sampling biases resulting from differences in capture probabilities among species can influence our impressions of relative abundance of different species in an area. Several studies have demonstrated that the effectiveness of traps in capturing small mammals differs with trap type, species, and geographic location (Cockrum 1947, Holdenried 1954, Kisiel 1972, McComb and others 1991, Slade and others 1993, Wiener and Smith 1972, Williams and Braun 1983). Fitch live-traps (John Rose, Norfolk, Virginia) present a potential alternative to Sherman live-traps (H.B. Sherman Traps, Inc., Tallahassee, Florida) . They are more economical, costing roughly 75% the cost of Sherman live-traps. Fitch traps are designed with a wire mesh entrance. O'Farrell and others (1994) found that live-traps made of mesh material caught some species of mammals not captured in Sherman live-traps and that capture success for some species was higher in mesh traps than in Sherman live-traps. In this paper, we compare the trap success of Fitch and Sherman live-traps in riparian areas of the Oregon Coast Range. STUDY AREA AND METHODS We selected 4 riparian areas in the Oregon Coast Range for study: Smith River (T21S R9W Sect. 5; Douglas County, elevation 250 m), Cleghorn Creek (T21S R7W Sect. 4; Douglas County, elevation 200 m), Honey Grove Creek (T14S R7W Sect. 3; Benton County, elevation 150 m)) and Bull Run (T13S R9W Sect. 14; Lincoln County, elevation 300 m). The sites are 25 to 60 km from the coast. Vegetation at Honey Grove Creek, Bull Run, and Cleghorn Creek was dominated by red alder (Alnus rubra) in the overstory with occasional bigleaf maple (Acer macrophyllum), Douglasfir (Pseudotsuga menziesii), western hemlock (Tsuga heterophylln), and western redcedar (Thuja pZicata). Shrub layers at these sites were dominated by salm-