Lithium clearance (C(Li)) was studied in the healthy subjects (H) and in patients with chronic renal failure (CRF) with a mean GFR of 36.08 +/- 3.8 ml/min x 1.73 m2 in order to evaluate (1) the variability of C(Li) measurements; (2) the changes of C(Li) during the hyperfiltration response to a meat meal (2 g/kg b.w. of protein), and (3) the effects on C(Li) Of different protein intakes. C(Li) correlated with GFR and with urinary sodium, Fe(Li) correlated with Fe(Na) in H and CRF. The intraindividual standard deviation (IISD) of GRF, C(Li) and Fe(Li) assessed in triplicate studies in the course of a single experiment averaged 4.70, 7.15 and 10.69% in CRF and 1.77, 4.09 and 4.61% in H. When IISD for GFR, C(Li), and Fe(Li) was assessed in triplicate measurements in the course of 3 studies performed at 1-week intervals (day 1, day 8, day 15), it averaged 7.36, 9.21 and 15.72% in CRF and 1.83, 5.08 and 5% in H. During the hyperfiltration response to a meat meal, C(Li) and Fe(Li) did not change in CRF and increased significantly in H. In 10 patients with CRF who were switched for 3 weeks from a protein intake of 0.85 g/kg to a protein intake of 0.60 g/kg, GFR was significantly (p < 0.001) reduced by 29%, while C(Li) was stable. The data indicate that because of its variability, C(Li) is of little use in experiments designed to study subtle changes in renal Na transport.