Skeletal muscle is formed by the iterative fusion of precursor cells (myocytes) into long multinuclear fibres. Extensive studies of fusion in Drosophila embryos have lead to a paradigm in which myoblasts are divided into two distinct subtypes - founder and fusion-competent myoblasts (FCMs) - that can fuse to each other, but not among themselves. Only founder cells can direct the formation of muscle fibres, while FCMs act as a cellular substrate. Recent studies in zebrafish and mice have demonstrated conservation of the molecules originally identified in Drosophila, but an important question remains: is vertebrate fusion regulated by specifying myocyte subtypes? Stated simply: do vertebrate founder cells exist? In light of recent findings, we argue that a different regulatory mechanism has evolved in vertebrates. Copyright © 2012 Elsevier Ltd. All rights reserved.