Axon growth inhibitory CNS matrix proteins, such as tenascin-R (TN-R), have been supposed to contribute to the poor regenerative capacity of adult mammalian CNS. With regard to TN-R function in low vertebrates capable of CNS regeneration, questions of particular interest concern the (co)evolution of ligand-receptor pairs and cellular response mechanisms associated with axon growth inhibition and oligodendrocyte differentiation. We address here these questions in a series of comparative in vivo and in vitro analyses using TN-R proteins purified from different vertebrates (from fish to human). Our studies provide strong evidence that unlike TN-R of higher vertebrates, fish TN-R proteins are not repellent for fish and less repellent for mammalian neurons and do not interfere with F3/contactin- and fibronectin-mediated mammalian cell adhesion and axon growth. However, axonal repulsion is induced in fish neurons by mammalian TN-R proteins, suggesting that the intracellular inhibitory machinery induced by TN-R-F3 interactions is already present during early vertebrate evolution. In contrast to TN-R-F3, TN-R-sulfatide interactions, mediating oligodendrocyte adhesion and differentiation, are highly conserved during vertebrate evolution. Our findings thus indicate the necessity of being cautious about extrapolations of the function of ligand-receptor pairs beyond a species border and, therefore, about the phylogenetic conservation of a molecular function at the cellular/tissue level.