Considerable controversy exists as to whether lithium maintenance therapy is associated with the development of renal insufficiency. In 1980 we initiated a prospective study of renal function in manic-depressive patients beginning lithium therapy. None of the patients had evidence of pre-existing renal disease. Sixty-five patients were entered, and 51 and 18 patients completed 1 and 3 years of follow-up, respectively. Lithium doses were titrated to the lowest level consistent with control of psychiatric symptoms; there were no episodes of overt lithium intoxication. Serum creatinine levels in all patients, and endogenous creatinine clearance in women, remained stable over the course of the study. In contrast, creatinine clearances (mean +/- SEM, ml/min/1.73 m2) in men significantly decreased over both 1 year (110 +/- 4 to 95 +/- 5, n = 21, p = 0.0126) and 3 years (107 +/- 4 to 80 +/- 11, n = 8, p = 0.0385) of evaluation. Although all patients demonstrated a mild reduction in renal concentrating ability after initiation of lithium, the decrease was not progressive over the course of the study. Quantitative urinary protein excretion did not change, and repeated urinalyses did not reveal any evidence of renal disease. Thus, lithium therapy appears to result in modestly reduced rates of glomerular filtration, as measured by endogenous creatinine clearance, in men receiving lithium maintenance therapy for manic-depressive illness. Whether this reduction is progressive and leads to clinically significant renal insufficiency requires further investigation.