Banksia (80 spp.; Proteaceae) has undergone extensive speciation and adaptive radiation on the island continent of Australia. Its members range from prostrate shrubs in the dry, infertile sandplains to 25 m tall trees in the loams of river margins, and they display striking variation in their fire survival strategies and floral and foliar morphologies. We examine the weight of both previously published (most trnL intron, trnL/F spacer, and rpl16 intron data) and new (matK, atpB, and waxy data, as well as most ITS data) DNA sequence evidence for the paraphyly of Banksia with respect to a monophyletic Dryandra (93 spp.). The nuclear waxy gene appears to be at two loci in the Proteaceae, and sequences presumably from the same locus resolve Banksia as paraphyletic with respect to Dryandra. The waxy and combined chloroplast DNA (cpDNA) data reject the monophyly of Banksia at a threshold of P = 0.05 using the winning sites and Kishino–Hasegawa tests. We consider this result and the repeated placement of Dryandra in the same clade (/Cryptostomata) of Banksia with each separate analysis of the DNA datasets (cpDNA, ITS, and waxy), to be strong molecular support for the paraphyly of Banksia with respect to Dryandra. The morphological synapomorphy of beaked follicles for /Cryptostomata (including Dryandra) reinforces this conclusion. We argue that realignment of taxa to produce one or more monophyletic genera is best attained by moving the taxa of Dryandra to Banksia. This would produce an easily recognised genus Banksia with four morphological synapomorphies. It would also probably confer some of the research attention garnered by the adaptive radiation of Banksia to the under-studied taxa of Dryandra, for Dryandra makes the radiation of Banksia even more remarkable.