Velvet ants are solitary wasps that have been virtually ignored in the past. Although over 4200 species have been described, fewer than 10 articles are published on these wasps each year. Their research potential is hindered by lack of phylogenetic and taxonomic data and lack of interest in the scientific community. In this dissertation, I sought to overcome the hindrances to mutillid research with a holistic systematic research model. By reconstructing phylogenies using molecular methods and correcting taxonomy based on the phylogenetic reconstructions, I was able to diminish the barriers to velvet ant research while concurrently presenting broadly interesting hypotheses. I applied this model to multiple hierarchical levels within the most widely studied velvet ant genus, Dasymutilla, and its allies. Molecular phylogenetic reconstructions, particularly those using the internal transcribed spacer units of ribosomal DNA (ITS1 and ITS2), were effective under Bayesian criteria. My results reveal the utility of velvet ants for studying biogeography and mimicry. I specifically determined that velvet ants dispersed between North and South America prior to the Great American Biotic Interchange and members of the genus Dasymutilla form the world’s largest known Müllerian mimicry complex. Taxonomic problems in Dasymutilla and their allies are addressed in the remaining sections of the dissertation. Overall, 35 species were treated taxonomically. Seven new sex associations were discovered, 22 species were recognized as synonyms, and nine new species were described. In each taxonomic treatment, hypotheses concerning biogeography, mimicry, and host selection were discussed briefly.