SUMMARY1. Intracellular activities of sodium, potassium and chloride ions, ai as, and at1were measured with ion-selective single-, double-and triple-barrelled micro-electrodes in skin and isolated epithelia of Rana temporaria bathed on both sides with normal or modified physiological saline. Apical and basolateral membrane potentials, Rfac and r,.s and resistance Ra and Rb respectively were also measured and from the latter the fractional resistance of the apical membrane, F(Ra) and voltage divider ratio, Aiac/Atcs were measured as criteria of satisfactory membrane penetration by the micro-electrodes.2. Under control conditions, aia was 12-3 + 0-8 mm, ai was 703 + 22 mm and at1was 20-3 + 1-6 mm with VRac averaging -38-0 + 3-2 mV. When 10-4 M-amiloride was added to the apical bathing fluid aia fell within 10 min to 118+001 mm and at1 to 52 + 09 mm, while a' increased to 86-2 + 3-8 mm as measured from the basolateral border of isolated epithelia.3. The sodium transport pool of the skin was measured from the fall in aia in the presence of amiloride and could be expressed as 33 x 10-9 mol cm-2 ofepithelium. The mean rate of fall of aia under these conditions corresponded to an efflux rate at the basolateral border of 30-1 x 10-9 mol cm-2 min-(48 ,sA cm-2) giving a half-time for turnover of the sodium transport pool of 33 s. 4. Reduction ofsodium concentration in the apical fluid from the normal 79 mM-Na to 10, 1 and 01 mm caused aka to fall in stages to 2 mm. Because Vfac increased in negativity to -101 mV in the process, this driving force for passive sodium accumulation, more than offset the increased sodium gradient opposing sodium influx across the apical border.