Soils are one of the most biodiverse and least understood terrestrial habitats. Arthropods likely comprise the most important component of soil fauna diversity yet are so poorly known that estimates of their richness span two orders of magnitude. Comprehensive assessments of the fauna have been impeded by the prevalence of undescribed species and the scarcity of taxonomic expertise, combined with their small size and often cryptic morphology. In this thesis, I employ DNA barcoding and metabarcoding to advance understanding of soil arthropods and one of their most diverse factions, the mites (Arachnida: Acariformes, Parasitiformes). I first utilize DNA barcodes to assess a hemi-continental fauna—the mites of Canada. This survey uncovered 2.4x the number of species previously recorded in Canada and indicates that the fauna likely includes more than 30,000 species. Mite assemblages showed high β-diversity and were spatially structured, reflecting dispersal-limited, environmentally driven assembly. This survey generated the most comprehensive reference library for any national mite fauna, with coverage for all four orders and 60% of the families known from Canada. I then utilized this library to evaluate the accuracy of higher-taxonomic assignments based on DNA barcodes in mites, work that made it possible to develop family and ordinal identification thresholds for both full-length DNA barcodes and for the partial sequences used in metabarcoding. Lineage-specific thresholds and an expanded reference library improved the success of higher-taxonomic assignments, and more strict thresholds were needed to reduce assignment errors from short sequences. Finally, I assessed the efficacy of metabarcoding two types of bulk samples (specimens, soil) for surveying the soil arthropod community. Despite recovering different fractions of the fauna, specimen and soil analyses revealed similar patterns of α- and β-diversity. Expanded soil analysis also confirmed patterns of highly dissimilar, spatially structured arthropod distributions, demonstrating the ability of metabarcoding to enable rapid, robust surveys of the fauna. However, deep sequence coverage will be necessary to develop comprehensive baselines for the soil fauna, and reference library expansion is needed to improve the taxonomic resolution for these surveys. This thesis demonstrates the power of DNA barcoding and metabarcoding to advance knowledge of arthropod diversity and the processes that foster it.