Determining how migratory animals are spatially connected between breeding and non‐breeding periods is essential for predicting the effects of environmental change and for developing optimal conservation strategies. Yet, despite recent advances in tracking technology, we lack comprehensive information on the spatial structure of migratory networks across a species’ range, particularly for small‐bodied, long‐distance migratory animals. We constructed a migratory network for a songbird and used network‐based metrics to characterize the spatial structure and prioritize regions for conservation. The network was constructed using year‐round movements derived from 133 archival light‐level geolocators attached to Tree Swallows (Tachycineta bicolor) originating from 12 breeding sites across their North American breeding range. From these breeding sites, we identified 10 autumn stopover nodes (regions) in North America, 13 non‐breeding nodes located around the Gulf of Mexico, Mexico, Florida, and the Caribbean, and 136 unique edges (migratory routes) connecting nodes. We found strong migratory connectivity between breeding and autumn stopover sites and moderate migratory connectivity between the breeding and non‐breeding sites. We identified three distinct “communities” of nodes that corresponded to western, central, and eastern North American flyways. Several regions were important for maintaining network connectivity, with South Florida and Louisiana as the top ranked non‐breeding nodes and the Midwest as the top ranked stopover node. We show that migratory songbird networks can have both a high degree of mixing between seasons yet still show regionally distinct migratory flyways. Such information will be crucial for accurately predicting factors that limit and regulate migratory songbirds throughout the annual cycle. Our study highlights how network‐based metrics can be valuable for identifying overall network structure and prioritizing specific regions within a network for conserving a wide variety of migratory animals.