To study proximal tubule bicarbonate absorption that is not due to the neutral Na+-H+ antiporter, mid to late proximal convolutions of the rat kidney were microperfused in vivo with a sodium-free choline solution containing 10(-3) M amiloride. The average sodium concentration resulting from sodium influx was 12 mM. At such low intraluminal [Na+], 10(-3) M amiloride should have inhibited the Na+-H+ antiporter by greater than 95%. When 25 mM HCO3- was in the perfusion fluid, measured total CO2 absorption was 100 pmol.mm-1.min-1. When luminal [HCO3-] was raised to 50 mM, and blood [HCO3-] was also raised to approximately 50 mM to avoid a transepithelial HCO3- concentration gradient, total CO2 absorption increased to greater than 300 pmol.mm-1.min-1. Thus raising intraluminal HCO3- concentration caused a marked increase in total CO2 absorption even though intraluminal [Na+] was low and amiloride was present. Control perfusions containing 140 mM Na+ yielded total CO2 absorption that was approximately 100 pmol.mm-1.min-1 higher than with the respective sodium-free perfusion solutions. In additional experiments, either DCCD or NEM was added to sodium-free perfusion solutions to inhibit H+-ATPase. These inhibitors reduced Na+-H+ independent total CO2 absorption markedly. Our observations suggest that under physiological acid-base conditions, sodium-independent H+ secretion can account for approximately 50% of total HCO3- absorption in mid to late proximal convolutions. This mechanism is stimulated by an increase in ambient HCO(-3) concentration to a degree that might account for the load-dependency of proximal HCO(-3) absorption in these segments of the proximal tubule.(ABSTRACT TRUNCATED AT 250 WORDS)