The hypotheses of passive salt accumulation predict an enhancement of renal concentrating ability by urea. We tested this prediction in rabbits, a species whose nephons when studied in vitro show tansport properties that support these hypotheses. We used calm, unanesthetized, hydropenic, vasopressin-treated rabbits with intact kidneys fed a 16% protein diet, and we observed the effect of urea administration at two rates of solute excretion (60 and 190 microOsm/min . kg body wt; N = 10 and 5, respectively). After an i.v. mannitol infusion, when urea was infused, the i.v. solute excretion rate was unchanged, the changes in urine urea concentration were large (a change of 767 and 408 mumoles/ml), but only small and variable changes in urine osmolality occured (a change of 78 +/- 146, and 36 +/- 50 microOsm/g H20). In additional experiments, we removed the kidneys from antidiuretic, or urea- or mannitol-infused rabbits and measured the intrarenal distribution of sodium, potassium, urea, and chloride. When the urine urea level was greater than 400 mmoles, the urine-to-papilla ratios for urea were 1.6 to 3.6. This suggested that a low collecting duct permeability to urea could explain the absence of a marked enhancement of concentrating ability during urea administration. Further analysis, based on a model of inner medullary solute compartments, indicated that sodium chloride was the major (86%) osmotically active solute in the medullary central core of these rabbits and that it was not influenced by changes in urinary urea concentration. The results of tissue analysis were consonant with either active or passive sodium chloride reabsorption from the thin ascending limb of Henle's loop in these rabbits.