The transepithelial transport of Na+ by amphibian skin must be accompanied by the corresponding anion, Cl-, and much effort has been devoted to the characterization of Cl- transport. The transepithelial Cl- conductance, G(Cl), is activated by voltage and adenosine 3',5'-cyclic monophosphate (cAMP), shows rectification, requires the presence of Cl- in the pathway and is influenced by factors modifying intracellular signalling cascades and by metabolic poisons such as cyanide (CN-). Until recently, these findings were interpreted as strong evidence for a transcellular path, for which, given the impermeability of the principal cells for Cl-, the mitochondria-rich cells (MRC) are the only candidate. This was supported by the apparent parallelism between G(Cl) and the density of MRC (D(mrc)). Data accumulated in recent years, however, raise serious doubts as to the validity of this concept. The single-channel conductance derived from various techniques is too small by an order of magnitude to account for the observed G(Cl), the very slow time course of conductance activation is not reconcilable with any known membrane channel gating processes, a more thorough examination of the relationship between G(Cl) and D(mrc) fails to show any consistent pattern and analysis of current density immediately above the transporting epithelium using the vibrating voltage probe shows current peaks associated with only a small fraction of MRC, and even so, these current peaks account for about 20% of the transepithelial current. The remaining 80% of the current cannot be localized to specific structures. Given the increasing evidence for close cellular control of tight-junction function, the foregoing findings are equally consistent with an additional, major, paracellular pathway for Cl-. A comprehensive description of Cl- transport must await the final resolution of the transport pathway(s).