Acidification in proximal tubule of the isolated rat kidney, perfused in vitro, was studied by stopped-flow microperfusion techniques, using Sb microelectrodes to measure luminal pH. The kidney was perfused with mammalian Ringer's solution at pH 7.4 buffered by 20 mmol/l phosphate and containing 7.5 g/100 ml bovine albumin, equilibrated with air. Final urine pH was 6.88 +/- 0.5. Steady-state pH in proximal segments was 6.81 +/- 0.03 (n = 80), and acidification half-time (t/2) 7.25 +/- 0.33 (80) s, giving a net secretory H+ ion flux of 0.51 +/- 0.05 nmol . cm-2 . s-1. This flux was about 70% of "in vivo" (blood perfused kidneys). During luminal perfusion with solutions at pH 6.2, back-flux of H+ was 0.82 +/- 0.08 nmol . cm-2 . s-1, with an alkalinization t/2 of 6.33 +/- 0.34 (34) s. The difference between acidification and alkalization t/2 was not significant. This is compatible with a pump-leak system of H+ transport. This is compatible with a pump-leak system of H+ transport. The back flux of H from the lumen was markedly reduced in low Na+ perfused kidneys in the presence of 10(-4) mol/l amiloride in the lumen, indicating that this process is mediated by the luminal Na/H exchanger. Observations in the presence of high K levels suggest that it may have also a charged component. 10(-4) mol/l acetazolamide added to the kidney perfusate reduced acidification to 0.5% of control, and 10(-6) mol/l SITS to 25% of control. Thus, despite the low pCO2 (0.1-0.4 kPa, or 1-3 mm Hg), the CO2/HCO-3 buffer system still plays an important role in tubular acidification in this preparation.