The effect of altering the volumes of different body fluid compartments on the renal response to atrial natriuretic peptide (ANP) was studied in anesthetized rats before and during administration of the peptide at 170 ng/min. Four different groups were used. In the first (De), reduction of total body water content was induced by 48 h water deprivation. In the second (De+NaCl), an acute intravenous infusion after the same 48 h dehydration was used to restore the extracellular, but not the intracellular, fluid compartment. In the third (Eu+NaCl), euvolemic rats were infused with isotonic saline at the same rate as in group De+NaCl to expand both intravascular and interstitial components of extracellular fluid. In the fourth group (Eu+BSA) an infusion of hyperoncotic (6%) bovine serum albumin in isotonic saline was used to expand the intravascular volume while contracting the interstitial volume. Excretion of water and salt was predictably reduced in the De group compared with the others. This reduction was associated with increased tubular reabsorption, both upstream from the medullary collecting duct and in the duct itself. Administration of ANP did not significantly affect diuresis and saluresis, or tubular transport. By contrast, there were marked and similar diuretic and natriuretic responses to ANP in groups De+NaCl and Eu+NaCl, associated with transport inhibition primarily in the medullary collecting duct. Surprisingly, the rats infused with hyperoncotic solution (Eu+NaCl) also failed to show marked excretory or duct transport responses to ANP. According to the study design, the two nonresponding groups had, respectively, a decreased or a normal intracellular compartment, and a decreased or increased plasma volume. The common feature of both nonresponding groups was a decreased interstitial fluid compartment, whereas the two responding groups had normal or increased interstitial volume. We suggest, therefore, that a replete interstitial fluid compartment is essential for the renal response to ANP.