The Entomophthoromycotina, a subphylum close to the root of terrestrial fungi with a bias toward insects as their primary hosts, has been notoriously difficult to categorize taxonomically for decades. Here, we reassess the phylogeny of this group based on conserved genes encoding ribosomal RNA and RNA polymerase II subunits, confirming their general monophyly, but challenging previously assumed taxonomic relationships within and between particular clades. Furthermore, for the prominent, partially human-pathogenic taxon Conidiobolus, a new type species C. coronatus is proposed in order to compensate for the unclear, presumably lost previous type species C. utriculosus Brefeld 1884. We also performed an exhaustive survey of the broad host spectrum of the Ento-mophthoromycotina, which is not restricted to insects alone, and investigated potential patterns of co-evolution across their megadiverse host range. Our results suggest multiple independent origins of parasitism within this subphylum and no apparent co-evolutionary events with any particular host lineage. However, Pterygota (i.e., winged insects) clearly constitute the most dominantly parasitized superordinate host group. This appears to be in accordance with an increased dispersal capacity mediated by the radiation of the Pterygota during insect evolution, which has likely greatly facilitated the spread, infection opportunities, and evolutionary divergence of the Entomophthoromycotina as well.