The tubuloglomerular feedback hypothesis states that filtration rate in a single nephron unit (SNGFR) is controlled or modified through variations in the rate of either distal delivery or tubular reabsorption in that same nephron. We have reexamined this hypothesis by varying the perfusion rate of the late proximal tubule from normal (10 nl/min) to supranormal values (30 nl/min) in a reversible fashion and measured the resultant nephron filtration rate. Three perfusion fluid were utilized: group I, a solution simulating late proximal tubular fluid (11 mM HCO3); group II, a solution with a higher HCO3 concentration (30 mM HCO3); and group III, a solution simulating early proximal tubular fluid. In group I SNGFR was 27.2 nl/min during perfusion from the late proximal tubule at 10 nl/min and fell to 19.8 nl/min during the 30 nl/min perfusion rate. Group II SNGFR was 27.7 nl/min at the 10 nl/min rate and 19.0 nl/min during the 30 nl/min perfusion. Groups I and II then demonstrated a completely reversible tubuloglomerular feedback response. In group III, the SNGFR at 10 nl/min was 26.2 nl/min and 23.4 nl/min at the 30 nl/min perfusion rate and were significantly different. Evaluation of collections of distal tubule fluid in perfused nephrons revealed that increased absolute reabsorption of water and chloride between late proximal and distal tubules was associated with the reductions in SNGFR in the same nephron. A strong candidate for the afferent limb of this system is the change in absolute chloride reabsorption between late proximal and distal tubules.