Estimations of proximal tubule sodium reabsorption with the FELi method come closer to direct measurements than any other indirect method. There is little doubt that most lithium reabsorption takes place in the proximal tubules, very likely in proportion to the reabsorption of sodium and water. It is also likely that changes in proximal tubule sodium reabsorption due to changes in volume status are paralleled by changes in proximal tubule lithium reabsorption, at least in the superficial nephrons. Nonetheless, changes in FELi probably do not purely reflect changes in proximal reabsorption, since lithium is also handled beyond the proximal tubules. Acknowledged problems are lithium reabsorption in Henle's loop and in the late distal and collecting tubules. The latter occurs in the rat and the dog, but not or much less in men. Sodium restriction enhances this lithium transport considerably. It is as yet uncertain whether other conditions, such as increased vasopressin activity or lowering of renal perfusion pressure, also influence this transport. Amiloride appears to prevent this reabsorption of lithium. Therefore, this drug can be used in lithium clearance studies whenever unwanted "distal" lithium reabsorption is expected. Lithium reabsorption in Henle's loop forms a greater problem as it cannot be prevented by any drug without influencing proximal tubule reabsorption. It is estimated that about 7% of the filtered lithium (one-tenth of total lithium reabsorption) is normally taken up here, preferentially in deep nephrons. In view of studies with furosemide, this reabsorption probably varies with sodium intake, but the proportion of this variation to that of proximal tubule lithium reabsorption is obscure. This remains an uncertain factor in any circumstance where the lithium clearance method is used. In some conditions the change in FELi may be so large relative to the expected changes in proximal reabsorption, that use of FELi as marker of end-proximal solute delivery seems unjustified. Disproportionately large suppression is likely during mineralo-corticoid-induced volume expansion, and stimulation during prostaglandin synthesis inhibition and vasopressin. Based on observations in these conditions the potential range of lithium reabsorption in the loop of Henle would be 0 to 15% of filtered load. In this review attention was paid mainly to the validity of lithium clearance as a pure "proximal marker". Many of our interpretations suffer from incomplete certainty with respect to the renal effects of tested maneuvers, a problem which is acknowledged.(ABSTRACT TRUNCATED AT 400 WORDS)