In order to determine whether HCO(3) (-) gains access to the proximal tubular lumen from a source other than the glomerular filtrate, we carried out microperfusion experiments on isolated segments of rat proximal tubules in vivo. The perfusion fluid was essentially free of HCO(3) (-) and of a composition that prevented net absorption of sodium and water.It was found that when plasma HCO(3) (-) concentration and CO(2) tension (PCO(2)) were normal, the HCO(3) (-) concentration in the collected perfusate rose to about 3 mEq per L. Inhibition of renal carbonic anhydrase did not produce an appreciable change in this value in normal rats, but when the enzyme was inhibited in acutely alkalotic rats, a mean concentration of 15 mEq per L was recovered in the perfusate. Addition of HCO(3) (-) to the tubular lumen might occur by either intraluminal generation of HCO(3) (-) from CO(2) and OH(-) or by influx of ionic bicarbonate from the plasma or tubular cells. Because of the marked increase in HCO(3) (-) found when intraluminal carbonic anhydrase was inhibited, generation of new HCO(3) (-) from CO(2) and OH(-) seems unlikely. We conclude, therefore, that influx of ionic bicarbonate occurred, either across the luminal membrane or through extracellular aqueous channels. These observations suggest that the proximal epithelium has a finite degree of permeability to HCO(3) (-) and that influx of this ion may be a component of the over-all handling of HCO(3) (-) by the kidney.