The nomadic Sedge Wren (Cistothorus platensis) breeds primarily in mesic grasslands in north-central North America. Following breeding in these regions from late May to early July, however, the species then “appears” en masse in the tallgrass prairie region farther south (e.g., Missouri and Kansas) and to the east to breed again from mid-July to early August (Herkert et al. 2001). The provenance of birds appearing in late summer to breed in these areas remains unknown because of problems inherent in mark—recapture surveys. Recent studies have shown how endogenous markers may be used to infer origins of individual birds. We analyzed levels of the stable hydrogen isotope ²H (δD) from liver, muscle, and claws of Sedge Wrens from known northern breeding locations first to establish the relationships between δD in the wrens' tissue and mean δD in precipitation during the growing season (δD_p ). From these relationships we derived expected values (mean and 95% CI) for three sites in Kansas and Missouri where late breeders colonized. The observed values of δD in these late breeders were primarily within the range expected for those locations, but more individuals than expected had δD values higher than expected. In addition, in birds apparently originating from north or south of Kansas and Missouri, the values of δD in claws were positively correlated with those in other tissues, in contrast to those with the “local” signal. This supports the idea that the isotopic outliers at these sites were more recent arrivals. For small-bodied birds like the Sedge Wren, however, the isotopic approach based on soft tissues is limited to a very narrow temporal window of inference because of rapid elemental turnover. This greatly restricts the use of this technique in inferring origins of small nomadic species.