The parasitoid lifestyle is largely regarded as a key innovation that contributed to the evolutionary success and extreme species richness of the order Hymenoptera. Understanding the phylogenetic history of hyperdiverse parasitoid groups is a fundamental step in elucidating the evolution of biological traits linked to parasitoidism. We used a genomic-scale dataset based on ultra-conserved elements and the most comprehensive taxon sampling to date to estimate the evolutionary relationships of Braconidae, the second largest family of Hymenoptera. Based on our results, we propose the existence of 41 braconid subfamilies, confirmed a number of subfamilial placements and proposed subfamily-level taxonomic changes, notably the restoration of Trachypetinae stat. rev. and Masoninae stat. rev. as subfamilies of Braconidae, confirmation that Apozyx penyai Mason belongs in Braconidae placed in the subfamily Apozyginae and the recognition of Ichneutinae sensu stricto and Proteropinae as non-cyclostome subfamilies robustly supported in a phylogenetic context. The correlation between koinobiosis with endoparasitoidism and idiobiosis with ectoparasitoidism, long thought to be an important aspect in parasitoid life history, was formally tested and confirmed in a phylogenetic framework. Using ancestral reconstruction methods based on both parsimony and maximum likelihood, we suggest that the ancestor of the braconoid complex was a koinobiont endoparasitoid, as was that of the cyclostome sensu lato clade. Our results also provide strong evidence for one transition from endo- to ectoparasitoidism and three reversals back to endoparasitoidism within the cyclostome sensu stricto lineage. Transitions of koino- and idiobiosis were identical to that inferred for endo- versus ectoparasitoidism, except with one additional reversal back to koinobiosis in the small subfamily Rhysipolinae.