1. To study the effects of chloride depletion, without sodium depletion or change in plasma tonicity, on renal excretion of sodium and water, a single exchange peritoneal dialysis was performed in rats against a solution of glucose (15 g/1) containing either NaCl (150 mmol/l, control) or NaHCO3 (150 mmol/l, experimental); KHCO3 (4mmol/l)was added to both solutions. All rats were prepared before dialysis by a low NaCl diet for 10 days. 2. Peritoneal dialysis against NAHCO3 consistently produced a negative sodium and water balance compared with dialysis against NaCl. Despite this, subsequent electrolyte balance for 3 days showed that chloride-depleted rats excreted significantly more sodium and water and had a reduced urinary osmolality as compared with control animals. Increased sodium and water loss were unexplained by osmotic or bicarbonate diuresis. Kaliuresis was seen in the chloride-depleted rats but muscle potassium was not significantly depressed. 3. With sodium and water loss and continued renal chloride conservation, plasma chloride rose on the average from 88 mmol/l after dialysis against NaHCO3 to 100 mmol/l (control 104 mmol/l) at 72 h. Concomitant with this increase in plasma [C1-], on the third day after dialysis, during hydropenia, urinary osmolality and papillary [Na+] were not different from control cencentrations. 4. It is postulated that chloride depletion and/or hypochloraemia leads to diminished chloride transport in the loop of Henle and that this causes reduced sodium transport into the medulla, impaired concentration ability and inappropriate urinary sodium loss.