To assess the intrinsic effects of treatment with furosemide on free-water excretion in patients with chronic renal failure, two groups of patients with and without replacement of diuretic-induced salt losses have been studied. Furosemide therapy was administered for 1 week during constant sodium intake (100 mEq/day). In neither of the groups did furosemide cause hyponatremia, while it did decrease the urine to plasma osmolality ratio, an effect lasting even when the diuretic effect was exhausted. During water diuresis, furosemide decreased the fractional sodium reabsorption in diluting segments but not the absolute rate of the free-water generation (CH<sub>2</sub>O). Presumably the expected decrease of CH<sub>2</sub>O was masked by the increased distal delivery of tubular fluid mainly due to an additional effect of the diuretic on the proximal tubule. The hypotonicity of urine after furosemide treatment may be secondary to the dissipation of medullary hypertonicity, caused by furosemide, in the condition of decreased water permeability of the collecting duct due to uremic disease.