Natural scientists in the nineteenth century were fascinated by the problem of the ancestry and origin of the vertebrates. Indeed, searching for ancestors, kinship, and relationships (systematics and classification) was as major a spur to Darwin’s life work as was his desire to understand how new forms arise. From around the mid-nineteenth century, evidence from adult and embryonic animals was used to support a large number of potential (and more or less likely, as it turns out) vertebrate ancestorsFannelids, echinoderms, hemichordates, and cephalochordates being some but by no means all. Different theories and theorists placed their emphasis on individual but different characters, some in embryos some in adults. Careers were built around the importance for vertebrate genealogy of a notochord in cephalochordates (Branchiostoma spp. and Asymmetron spp., commonly known as amphioxus) or of ciliary bands on ascidian larvae. By the mid-twentieth century, analyses based on sets of characters placed the cephalochordates as the closest relatives (sister taxon) to vertebrates. Consequently (i) the neural crest (NC)Fa synapomorphy of vertebratesFwas presumed to have arisen in a cephalochordate-like ancestor, and (ii) evidence for the antecedent of the NC (a protoneural crest) was sought in the extant cephalochordate, amphioxus, which was regarded as a surrogate for the prevertebrate ancestor. The presence in amphioxus embryos of migratory cells associated with the crests of the dorsal neural tube, and the expression in those cells of genes also expressed in vertebrate neural crest cells, opened a window onto the prevertebrate condition. Extant cephalochordates lack pigment cells, which are a NC-derivative in vertebrates. Therefore, the ancestral cell type associated with a protoneural crest has been regarded widely as a neural cell. Although ascidians were not regarded as the closest relatives to vertebrates, their cells and genes have been explored for signs of a protoneural crest. Such studies have acquired a greatly increased importance with the publication of several molecular phylogenies that place ascidians rather than cephalochordates as the sister group to vertebrates. In a recent paper on chordate ancestry of the neural crest, Jeffery (2007) laid out the consequences of these revised phylogenetic relationships. The gene Snail1 illustrates how the new phylogeny requires us to take a new approach to the origin of the NC. Snail1, an important marker for NC cells in vertebrates, is not expressed in mesoderm. AmphiSnail1 (i.e., Snail1 in amphioxus) is expressed in presomitic mesoderm. If the prevertebrate was cephalochordate-like, the dorsal neural tube (NC) expression of Snail1 was enhanced and co-opted from mesoderm, and the phylogenetically older mesodermal expression was lost with the origin of the vertebrates. Snail1-positive cells also are expressed in ascidians. In the sea vase, Ciona intestinalis, these cells form the ependyma, a layer of cells of neural ectodermal origin that line the spinal cord of the central nervous system. In ascidians, therefore, and unlike cephalochordates, Snail1-positive cells are neuronal not mesodermal. Furthermore, the primary function of migratory neural-tube-derived cells in ascidians is as pigment cell precursors; extant cephalochordates lack pigment cells. Ascidians possess a band of pigmented cells along the dorsal aspect of the neural tube. These cells arise in the neural tube fromHNK-1and Zic3-positive cells (HNK-1 is a NCmarker and Zic3 is required for neural induction in vertebrates) that then migrate into the body wall. Jeffery (2007) does not take the ‘‘pre-adaptive’’ view of identifying pigment cell precursors in ascidians as evidence of protoneural crest cells. Rather, he identifies them as an evolved lineage in their own rightF‘‘neural crest-like cells,’’ whose primary role is to provide pigment cells. With ascidians as the sister group to vertebrates, and neural crest-like cells as a lineage established and functioning in ascidians before the origin of NC cells in vertebrates, we have to reevaluate the data on gene expression patterns in amphioxus; reconsider whether the neural crest-like cells in ascidians functioned as a protoneural crest that co-opted genes and gene pathways from mesoderm, non-neural ectoderm or pharyngeal endoderm as basal cephalochordates are proposed to have done (Meulemans and Bronner-Fraser EVOLUTION & DEVELOPMENT 10:3, 261 –262 (2008)