Greatly reduced floral structures and a propensity for phenotypic plasticity in this group have long hampered phylogenetic hypotheses and circumscription of taxa on the basis of morphology alone. This study explores molecular and anatomical data for their utility in estimating the phylogeny of Phoradendron. Consistent with other parasitic plants, Phoradendron shows accelerated molecular sequence evolution. Sequence data from the nuclear ribosomal (nr) ITS region provide resolution among closely related species, but alignment of sequences from more divergent taxa and especially outgroups is problematic. Sequences comprising expansion domains D2 and D7 of the nr 26S gene are more conserved. Using a parsimony model of evolution, combined analysis of the three regions of the nr cistron yields stronger statistical support, as evidenced by bootstrap values, than each region analyzed separately. Although sampling of Phoradendron needs to be extended, particularly to South America, preliminary analyses have already provided answers to several questions, including the non-monophyly of conifer parasitism and the distant relationship of P. californicum to other North American species. Congruence and incongruence between molecular, anatomical and morphological data are examined.

Key words: 26S, parasitic plants, Phoradendron, phylogenetics, Santalales, Viscaceae