Rafinesque's big-eared bat, Corynorhinus rafinesquii, is considered rare and/or declining throughout its range. Concrete bridges are potentially important roosts for C. rafinesquii, especially in the Gulf Coastal Plain where the species' natural roosts (caves and large hollow trees) are inherently scarce. Successful efforts to monitor and conserve this species must account for its movements among multiple roosts and determine the duration of its roost use (including bridges) at different temporal scales. Therefore, I investigated roosting ecology of C. rafinesquii from 2000-2005 within a mixed hardwood-pine ( Pinus spp.) system in southeastern Mississippi. I conducted surveys of concrete bridges to determine phenological pattern of use and found that maternity colonies began to arrive at bridges as early as 9 March (in 2000), increased in size and abundance as spring progressed (with pups being born in mid-to late May), and persisted through August (with pups nursing as late as 25 July [2001]). Solitary C. rafinesquii roosted under bridges throughout the year, but general scarcity of bats found under bridges during cooler months implied use of alternate roosts. To locate such structures, I captured and radiotagged 25 C. rafinesquii at bridge roosts and subsequently attempted to find these individuals. Radiotagged bats used 14 hollow trees (Nyssa spp. and Magnolia grandiflora ) and 11 human-made structures (e.g., bridges, abandoned houses) as roosts. Radiotagged bats switched roosts every 2.1 days, switched roosts 2.6 ± 2.0 (mean ± SD) times and used 2.5 ± 1.2 roosts per tracking period (9.1 ± 2.6 days). Bats showed low daily fidelity to tree roosts, which were relatively common in some areas (but not exceptionally stable), and maintained higher fidelity to human-made roosts that were rare but of higher structural integrity. To examine roost fidelity of C. rafinesquii over longer time periods, I relied on recoveries of banded bats at bridges. Of 144 bats captured and banded, I recaptured 55; age-class (juvenile vs. adult) affected probability of recapture. In most instances (91 percent) recaptured bats were found at their original roost. Distance that a marked bat had moved from its initial roost (0-4 km) did not correlate with the length of time from its banding to its first (or only) recapture, indicating that C. rafinesquii maintained long-term fidelity to bridges, up to 4 years by some individuals. Results of this investigation corroborate that C. rafinesquii possesses low vagility and likely perceives its environment at a fine-grain scale. Thus, loss of its habitat on even a local level could have deleterious effects.