Knowledge of the voltage dependencies of apical and basolateral conductances is important in determining the factors that regulate transcellular transport. To gain this knowledge it is necessary to distinguish between cellular and paracellular currents and conductances. This is generally done by sequentially measuring transepithelial current/voltage (It/Vt) and conductance/voltage (gt/Vt) relationships before and after the abolition of cellular sodium transport with amiloride. Often, however, there are variable time-dependent and voltage-dependent responses to voltage perturbation both in the absence and presence of amiloride, pointing to effects on the paracellular pathway. We have here investigated these phenomena systematically and found that the difficulties were significantly lessened by the use of an intermittent technique, measuring It and gt before and after brief (less than 10 sec) exposure to amiloride at each setting of Vt. I/V relationships were characterized by these means in frog skins (Rana pipiens, Northern variety, and Rana temporaria). Cellular current, Ic, decreased with hyperpolarization (larger serosa positive clamps) of Vt. Derived Ic/Vt relationships between Vt = 0 and 175 mV (serosa positive) were slightly concave upwards. Because values of cell conductance, gc, remained finite, it was possible to demonstrate reversal of Ic. Values of the reversal potential Vr averaged 156 +/- 14 (SD, n = 18) mV. Simultaneous microelectrode measurements permitted also the calculation of apical and basolateral conductances, ga and gb. The apical conductance decreased monotonically with increasing positivity of Vt (and Va). In contrast, in the range in which the basolateral conductance could be evaluated adequately (Vt less than 125 mV), gb increased with more positive values of Vt (and Vb). That is, there was an inverse relation between gb and cellular current at the quasi-steady state, 10-30 sec after the transepithelial voltage step.