1. K influx from peritubular space into renal tubular cells, varphi(i) (K), was measured in doubly perfused Necturus kidneys by studying tissue uptake of (42)K added exclusively to the portal circulation. Concomitantly, net tubular Na reabsorption, varphi(n) (Na), was measured by clearance techniques. varphi(n) (Na) and varphi(i) (K) were varied widely by replacing solutions of physiological composition (controls) with solutions containing high K, low K, low Na, cyclamate instead of Cl, ouabain (10(-7)-10(-4)M) or ethacrynic acid (10(-5)-10(-4)M).2. The ratio of varphi(n) (Na) to varphi(i) (K) was found to vary with the experimental conditions, the control value of about 2 was maintained over a threefold variation in absolute Na reabsorption. This ratio increased with low K or ouabain to values near 4. With high K, ethacrynic acid, low Na or cyclamate the relationship was one or lower. Thus, net Na reabsorption can be uncoupled from peritubular K influx.3. These results can be best explained if there are two Na pumps working in parallel: pump A transporting Na (with Cl) and pump B, a Na-for-K-exchange pump. The ratio of Na efflux to K influx could approach infinity if only pump A works (if B is inhibited) and could approach one if only B works. It should vary between these limits in controls when both pumps are active, or when neither of the two pumps is completely inhibited.4. Alternatively, the experimental findings could be explained by a Na pump with a coupling ratio that varies within two extreme values, from high Na-K ratios (with Na reabsorption at, or near, control values but with very low K influx values) to low ratios (with normal K influx values but with low Na reabsorption values).