The involvement of potassium in transepithelial sodium transport was tested by studying net potassium reuptake by potassium-depleted frog skin epidermis. Normal potassium content in half-strength Ringer's (0.244 μequiv/mg dry weight) fell 43% after 16 h in K-free medium at 5°C. Reaccumulation, against an electrochemical potential gradient, to 83% of the initial tissue potassium content occurred following incubation for 4 h at 22°C in K-containing medium. Sodium was required in the solution bathing the inside, but not the outside surface of the skin, for net potassium reaccumulation. Ouabain caused an additional potassium loss from potassium-depleted epidermis, but did not have the same effect on potassium-depleted isolatedcells. Procaine, lithium and caffeine completely inhibited, antidiuretic hormone and cyclic AMP may partially inhibit and amiloride had no effect on potassium reaccumulation. In many cases decreases in sodium and water content were found to occur even in the absence of net potassium reaccumulation. The results suggest (1) potassium is actively transported into the epidermis, (2) this transport is not rigidly coupled to sodium extrusion or water loss, (3) potassium uptake is not rigidly coupled to transepithelial sodium transport, or only a small fraction is involved, (4) potassium diffusion is restricted in the extracellular space.