Using serosal exposure of the isolated frog skin to collagenase followed by hydrostatic pressure we isolated sheets of epithelia with differing number of intact submucosal glands in situ (Andersen et al., 1995). The epithelial sheet was mounted in a chamber with the serosal side up and glands were patched using the Nystatin (slow) whole-cell method with a high-K⁺ pipette solution (126 mM K⁺, 42 mM Cl^-) Resting cells had zero-current potentials of -64.3±1.7 mV (mean±SEM,n=32) The zero-current potential was mainly governed by a K⁺-equilibrium potential, shown by a linear dependence of the potential on log(bath-K⁺) with a slope of 31 mV/decade (bath=3.7, 10, 25 mM K⁺, respectively, K⁺ substituting Na⁺) A population of cells had currents that activated with depolarization to positive potentials giving a highly outwardly rectifying I/V-relationship. The activating currents were blocked by 5 mM Ba²⁺. Stimulating cells with 10^-5 M epinephrine gave nse to a depolansation of 34.5±2.9 mV (n=10), which was mimicked by 2×10^-6 M isoproterenol (depolarisation: 34.5±3.4 mV, n=8). The depolansation was sustained or slowly inactivating in the presence of the agonist and reversible upon withdrawal. On stimulated cells, addition of 10⁴ M amiloride led to a highly variable reversible hyperpolarization (range 0-18 mV, mean = 5 8±1.9 mV, n=9) dependent on bath-Na⁺. This supports the recirculation hypothesis for isotonic secretion (Ussing et al , 1996), stating that apical Na⁺ channels are activated in parallel with Cl^- channels to recirculate part of the secreted Na⁺ in order to achieve isotonicity of the secretion.