Small lipid-soluble cations, such as tetraphenylphosphonium (TPP+) and tetraphenylarsonium (TPA+) are frequently used as probes of membrane voltage (delta psi, or Vm) for small animal cells, organelles, and vesicles. Because much controversy has accompanied corresponding measurements on 'walled' eukaryotic cells (plants, fungi), we studied their transport and relation to Vm in the large-celled fungus Neurospora crassa-where Vm can readily be determined with microelectrodes-as well as in the most commonly used model eukaryotic cell, the yeast Saccharomyces cerevisiae. We found no reasonable conditions under which the distribution of TPP+ or TPA+, between the cytoplasm (i) and extracellular solution (o), can serve to estimate Vm, even roughly, in either of these organisms. When applied at probe concentrations (i.e., < or = 100 microM, which did not depolarize the cells nor deplete ATP), TPP+ stabilized at ratios (i/o) below 30 in both organisms. That would imply apparent Vm values positive to -90 mV, in the face of directly measured Vm values (in Neurospora) negative to -180 mV. When applied at moderate or high concentrations (1-30 mM), TPP+ and TPA+ induced several phases of depolarization and changes of membrane resistance (Rm), as well as depletion of cytoplasmic energy stores. Only the first phase depolarization, occurring within the perfusion-turnover time and accompanied by a nearly proportionate decline of Rm, could have resulted from TPP+ or TPA+ currents per se. And the implied currents were small. Repeated testing, furthermore, greatly reduced the depolarizing effects of these lipid-soluble ions, implicating an active cellular response to decrease membrane permeability.