We sampled small mammals in burned and unburned tallgrass prairie by using three sites in each of two contiguous ungrazed experimental fire-reversal treatments during 1999-2010. One of the experimental treatments (001A to R20A) had been burned in spring for >20 years and then was switched to an unburned research treatment after the spring fire in 2000. In contrast, the other treatment (020A to R01A) had been left unburned for 20 of 29 years before it was switched to an annually burned treatment beginning with the prescribed fire in spring 2001. Overall, we recorded 11 species of rodents and two of shrews for a total of 2,444 individuals. The white-footed mouse (Peromyscus leucopus) was the most abundant species followed by the hispid cotton rat (Sigmodon hispidus) and deer mouse (P. maniculatus). Of the seven common species (>25 individuals), six were strongly associated with one of the treatments, and several species showed positive associations with some part of the landscape within each treatment. Furthermore, white-footed mice, deer mice, prairie voles (Microtus ochrogaster), western harvest mice (Reithrodontomys megalotis) and Elliot's short-tailed shrews (Blarina hylophaga) showed significant temporal patterns of abundance in either R01A or R20A in one or both seasons (autumn or spring). Our study also supports a very important conservation message. That is, the white-footed mouse (a woodland species) and the hispid cotton rat (a colonizing species) were more numerically dominant than the most common rodent, the deer mouse, (a native prairie species) throughout the 10-year study. These observations demonstrate that areas not burned frequently allow encroachment by shrubs and then trees, which subsequently allow the white-footed mouse and hispid cotton rat to expand into these areas. Conversely, the stoppage of frequent fires ultimately degrades the mosaic of prairie habitats for native prairie small mammals [such as the deer mouse, prairie vole and western and plains harvest mice (R. montanus)] and these species move out as shrubs and trees become common. Visual observations indicate that the latter (degradation of native tallgrass prairie) occurs much more rapidly (potentially within a decade) than the restoration of native tallgrass prairie by the elimination of shrubs and trees by annual burning (potentially 50 or more years).