Churchill, Manitoba, Canada is a diverse Subarctic region that has been the focus of DNA barcoding efforts and research for decades. Despite this effort, there are still taxa in this region that are underrepresented in public databases, particularly hyperdiverse insect groups. We present a collaborative large-scale DNA barcode reference library for Diptera based on molecular data and expert taxonomic identification. The reference library contains 16786 specimens sampled around Churchill from 2005 to 2011. Specimens were identified by morphological means, sequenced for the animal barcode marker, and sorted into 2235 Molecular Operational Taxonomic Units (using Barcode Index Numbers: BINs) representing 68 families and 1212 species. Eighty-six species shared a BIN with another species, 98 species were split across several BINs, and 1027 species were assigned to a unique BIN. The intraspecific and nearest neighbour distances varied across families, but most species can be easily distinguished from their nearest neighbours. By combining molecular data with taxonomic expertise, the barcode library provides species-level information for 12612 specimens, supplying future researchers with detailed taxonomic information and the opportunity to perform barcode-based specimen identifications for large-scale studies through sequence matching. Comparison of the Diptera composition of Churchill to other geographic regions reflected likely postglacial colonization northwards as well as a Beringian component. This study provides a publicly available, detailed reference library for use in future research, as well as new insight into the current diversity of Diptera present among the diverse habitats in this Subarctic region.