Since the early twentieth century, with the interest from botany greats such as Eastwood, Arctostaphylos (manzanitas) have challenged and captivated biologists. Although the majority of experts who have worked with the genus conclude that manzanitas are extremely diverse, the precise number of manzanita species and phylogenetic relationships among those species is both an historical and contemporary topic of debate. With its current taxonomic treatment, manzanitas form the most diverse woody genus within the California Floristic Province, and the majority of this diversity is represented by narrow endemics of conservation concern. Manzanitas are often considered difficult to identify in part because authoritative taxonomic keys reference characters unique to the genus and/or require characters from multiple phenological stages. This, compounded with limited phylogenetic resolution using traditional molecular systematics methods, poses significant challenges for conservation biologists and land managers who need to differentiate rare manzanitas from widespread ones that often occur in sympatry.  Herein, I present a maximum likelihood ddRADseq phylogeny for 43 manzanita species (55 taxa including subspecies) and 3 outgroup taxa. This represents >70% coverage of currently recognized species, and taxa have been replicated to test for monophyly. The phylogeny has a high degree of resolution and the majority of narrow endemic species are supported as monophyletic. The phylogeny, however, also reveals high complexity apparently associated with hybridization. In addition to the two major clades recovered in nuclear ITS phylogenies, this study found a third early diverging clade. Here I present why this clade as well as numerous independent lineages recovered for certain taxa suggest evidence of widespread hybridization.