Bicarbonate reabsorption in the thick ascending limb of Henle's loop was examined by studies of free-water clearance (CH2O) and free-water reabsorption (TcH2O). During maximal water diuresis in the dog, CH2O/GFR was taken as an indes of sodium reabsorption in, and urine flow (V/GFR) as an index of delivery of filtrate to, this scarbonate, infusion of a nonreabsorbable solute (hypotonic mannitol) and administration of an inhibitor of bicarbonate reabsorption (acetaent, but less than that achieved with hypotonic saline infusion. This suggests that sodium that sodium bicarbonate is not reabsorbed in the ascending limb. Rather, it is the sodium chloride, swept out of the proximal tubule by osmotic diuresis due to nonreabsorbed mannitol or sodium bicarbonate, that is reabsorbed in the ascending limb thereby increasing CH2O, whereas the nonreabsorption of mannitol and sodium bicarbonate results in a depressed CH20 per unit V when compared with hypotonic saline. V/GFR is not a satisfactory index of delivery to the ascending limb during osmotic diuresis, since it includes water obligated by nonreabsorbable solutes. When a better index of delivery, the sum of the clearances of chloride (CC1) and free-water (CH2O) is used, hypotonic bicarbonate infusion, hypotonic mannitol infusion and acetazolamide administration increase CH2O/GFR per unit delivery to the same extent as odes hypotonic saline infusion. Studies in dogs and rats on TcH2O also indicate that sodium bicarbonate is an impermeant solute in the ascending limb. Osmotic diuresis due to sodium bicarbonate diuresis, produced either by inhibition of sodium bicarbonate reabsorption (acetazolamide, L-lysine mono-hydrochloride) or infusion of sodium bicarbonate, or mannitol diuresis both produced marked chloruresis and increased TcH2O to the same extent as did hypertonic saline infusion. If chloride excretion was almost eliminated by hemodialysis against a chloride-free dialysate (dogs) or prolonged feeding of a salt-free diet (rats), TcH2O formation was unimpaired if hypertonic saline was infused but virtually obliterated during mannitol or sodium bicarbonate diuresis. Sodium reabsorption in the ascending limb, therefore, appears to be dependent upon chloride as the accompanying anion. At any given rate of bicarbonate excretion, more cloride is delivered out of the proximal tubule (as estimated from CC1 + CH2O) with hypotonic sodium bicarbonate infusion than with acetazolamide administration. This suggests that magnitude of the chlorutesis accompanying bicarbonate diuresis depends, not only on osmotic diuresis due to nonreabsorbed sodium bicarbonate, but also on the extent to which concomitant changes in effective extracellular volume influence overall sodium chloride reabsorption.