A hypothesis of relationships among subfamilies, tribes, genera, and species of Polygyridae was established by Ken Emberton in 1995, using shell, behavioral, allozyme, and soft anatomical characters. We tested this hypothesis using four mitochondrial and two nuclear loci. We present data from 418 polygyrid individuals sequenced for one to six loci, including 110 named species (out of 294 nominal taxa) from 21 of the 24 recognized genera. We carried out phylogenetic and DNA barcoding analyses to examine relationships at the family, genus, and species-level. In our analyses, the subfamilies are not supported as monophyletic groups. The tribes Mesodontini, Ashmunellini and Vespericolini were recovered as monophyletic, while all other tribes were paraphyletic. Regardless of analysis method, we found a close, well-supported relationship between Mesodontini and Triodopsini, two tribes that were distantly related in Emberton's hypothesis. Most genera were recovered as monophyletic with the notable exceptions of Cryptomastix Pilsbry, 1839, Mesodon Rafinesque in Férussac, 1821, and Neohelix von Ihering, 1892. Of the species for which we had multiple individuals, populations of 27 formed monophyletic groups on our phylogenies, while 47 did not, indicating an urgent need for revisionary taxonomy at all levels of classification in this family.