Three species of coastal sparrows (Ammodramus nelsoni, Nelson's Sparrow; A. caudacutus, Saltmarsh Sparrow; and A. maritimus, Seaside Sparrow) reside in North Carolina salt marshes in winter during their nonbreeding periods. We analyzed the timing of migration, survivorship, and site fidelity of these species with mark—recapture data from five winters (2006–2010). By determining the percentage of individuals captured more than once, we documented a shift from transient to settled individuals during and following migration and used linear regressions to test whether this shift was related to mean monthly minimum temperatures. We also used MARK and capture histories over five winters to estimate probabilities of apparent survival and capture. On the basis of recapture data, the three species' populations were composed largely of transient individuals from October until the sparrows settled for the winter by late November. Our recapture data indicate that when coastal sparrows settle for the winter or return to the region in a subsequent year, they do so with high site fidelity. We found that the percent of Nelson's and Seaside sparrows captured more than once had a negative relationship with mean monthly minimum temperature. We estimated survival of Nelson's, Saltmarsh, and Seaside sparrows at 0.673, 0.520, and 0.483, respectively. Our data indicate that survival of these coastal sparrows through the nonbreeding season is comparable to existing estimates for their breeding populations. Surveys in the nonbreeding season and management plans should take into account the likely presence of a large proportion of transient individuals during migration.