Small mammals can be useful indicators of sustainability in terrestrial ecosystems; hence, research and inventory of populations and communities have increased dramatically in recent years. Sampling methodologies to measure abundance and diversity attributes include removal- (snap and pitfall traps) and live-trapping, with the former predominating. We tested the hypothesis that removal-trapping of small mammals would alter patterns of abundance and species diversity compared with control (nonremoval) sites. Small-mammal communities were intensively sampled in coastal coniferous forest habitat in southern British Columbia, Canada. In a pulse-removal experiment, mean abundance of Oregon voles (Microtus oregoni) was significantly lower in removal than in control sites, whereas abundance of shrews (Sorex spp.) and species diversity were significantly higher in removal than in control sites. Abundances of 3 uncommon species, the long-tailed vole (M. Iongicaudus), southern red-backed vole (Clethrionomys gapperi), and American shrew-mole (Neurotrichus gibbsii), were significantly higher in removal than in control sites. Our results indicate that removal-trapping can disrupt small-mammal populations and yield spurious values for community characteristics. Ethical concerns notwithstanding, ecological studies of small mammals should use live-trapping to yield accurate estimates of population and diversity attributes.