The northern boundary of the endangered Southwestern Willow Flycatcher (Empidonax traillii extimus) is currently approximated as running through southern Colorado and Utah, but the exact placement is uncertain because this subspecies shares a border with the more northern and non-endangered E. t. adastus. To help resolve this issue, we evaluated the geographic distribution of mitochondrial and nuclear DNA by sampling breeding sites across the four-corner states (Arizona, Colorado, New Mexico, and Utah). We found that breeding sites clustered into two major groups generally consistent with the currently designated boundary, with the exception of three sites situated along the current boundary. However, delineating a precise boundary that would separate the two subspecies is made difficult because (1) we found evidence for a region of intergradation along the boundary area, suggesting the boundary is not discreet, and (2) the boundary region is sparsely populated, with too few extant breeding populations to precisely locate a boundary. The boundary region encompasses an area where elevation changes markedly over relatively short distances, with low elevation deserts to the south and more mesic, higher elevation habitats to the north. We hypothesized that latitudinal and elevational differences and their concomitant ecological effects could form an ecological barrier that inhibited gene flow between the subspecies, forming the basis for the subspecies boundary. We modeled changes in geographic patterns of genetic markers as a function of latitude and elevation finding significant support for this relationship. The model was brought into a GIS environment to create multiple subspecies boundaries, with the strength of each predicted boundary evaluated on the basis of how much genetic variation it explained. The candidate boundary that accounted for the most genetic variation was situated generally near the currently recognized subspecies boundary, but should be more biologically meaningful because it incorporates the landscape features that may be driving separation of the subspecies. Even so, we caution that using any narrow boundary line as an indicator of subspecies identity could be misleading because biologically the boundary is a region of intergradation rather than a discrete line. Designating, a boundary ultimately becomes a regulatory and management decision based on how much of the genetic variation unique to a subspecies should be protected. We discuss how the results of this study can help guide this decision process by wildlife policy makers.