Cirsium(i.e.,“thistles”)are one of the most taxonomically challenging groups of Compositae in North America.This study represents the first attempt to infer a broadly sampled phylogeny of Cirsium in North America.The two main objectives are to:(i)test whether currently hypothesized species variety complexes(C.arizonicum,C.clavatum,C.eatonii,and C.scariosum)constitute monophyletic lineages,and(ii)recircumscribe any taxa that are identified as problematic.Phylogeny reconstructions based on DNA sequence data from two nuclear ribosomal regions and four plastid markers were used to infer evolutionary lineages and test species'delimitations.Eight species varietal complexes were resolved as polyphyletic.We recircumscribed these complexes and in doing so found evidence to support the recognition of six new taxa.We hypothesize that the extensive taxonomic difficulty within Cirsium is the result of several factors:(i)previously undescribed taxa,(ii)inadequate representation of taxa from herbarium specimens,(iii)phenotypic convergence,(iv)hybridization,and(v)incipient speciation.While we can provide evidence to support the recircumscription of some taxa,others remain unresolved.Therefore,we are working on a phylogenomic study of North American Cirsium to answer remaining taxonomic questions as well as provide a robust framework for biogeographic studies.