SUMMARY1. Intracellular Na, K and Cl activities (ac8, ac and ac1) and membrane potentials were measured in Necturus gall-bladder epithelium using double-barrelled ion-sensitive micro-electrodes. Mucosal membrane potential was about -55 mV and the mean control activities were aca = 14-7 mm, ac = 91-6 mm and ac1 = 20-3 mM.2. Replacing mucosal Na by K caused a fall in aca that followed an exponential time course. The rate of change in aea was linearly related to aca above a certain value (-3 mm). ac and ac, both increased in K Ringer solution. From the change in all three ions the cell was estimated to swell at an initial rate of 0-13 % s-1. From the initial rate of change in aca, a net cell efflux of Na of 405 pmol cm-2 s-1 was calculated. Replacement of Na by Tris or choline led to a similar result. The transepithelial Na transport rate was for this group of animals 346 pmol cm2 s .3. Ouabain (10-3 M) produced an increase in aca and ac1, whereas ac decreased. The cells were estimated to swell at an initial rate of 0-06 % s-1. The initial Na influx after Na-pump inhibition was calculated to be 162 pmol cm-2 s-1. The parallel measure ofthe transepithelial rate oftransport ofNa gave a value of 189 pmol cm-2 s-1. Ouabain inhibited the decrease in aca after replacement of Na by K by about 80 %. 4. A fast depolarization, ranging from 2 to 7 mV, occurred after the perfusion with ouabain. Em then slowly decreased from about 53 to 32 mV in 1 h. 5. It is concluded that (a) the major fraction of the transepithelial transport of Na is transcellular and mediated by the Na pump, (b) the pumping rate is linearly dependent on internal Na within a certain range and (c) the Na pump is electrogenic under normal circumstances.