The effect of lithium on thirst and plasma vasopressin concentration was tested in seven subjects with affective psychiatric disorders. Mean ad libitum fluid intake was liberal but no different before (3,293 ml/day) and three to four weeks after treatment with lithium (3,443 ml/day). After fluid deprivation, plasma vasopressin was 1.5 ± 0.39 pg/ml before and 3.72 ± 0.55 pg/ml after treatment with lithium (p < 0.02) as plasma osmolalities and body weights were comparable. Urinary osmolalities were no different (735 versus 759 mOsm/kg) and did riot increase with exogenous vasopressin. With a water load, plasma vasopressin decreased 1.58 to 0.79 pg/ml (p < 0.05) before and from 2.68 to 0.91 pg/ml (p < 0.025) after treatment with lithium. The water load excreted in 4 hours was less during lithium therapy (66 versus 85 per cent, p < 0.05). Lithium therapy had no effect on plasma renin activity (PRA) or aldosterone. On standing, PRA increased from 2.27 to 5.28 ng/ml/hour (p < 0.05) before and from 2.19 to 7.59 ng/ ml/hour (p < 0.05) after lithium therapy. At the same time plasma aldosterone increased from 121 to 365 pg/ml (p < 0.05) before and from 76 to 436 pg/ml (p < 0.05) after treatment with lithium. Lithium had no effect on indices of proximal tubular function (HCO 3 −, HPO 4 =, glucose, amino acid and uric acid excretion). A lower titratable acid excretion (21 ± 5 versus 32 ± 4 μeq/min, p < 0.05) and higher urine pH (5.40 versus 5.02, p < 0.05) was observed after NH 4Cl ingestion during lithium therapy as compared to control. In conclusion, three to four weeks of lithium therapy neither stimulates thirst nor suppresses vasopressin release; some of the polyuria in patients with affective disorders may be due to their liberal fluid intakes. Lithium does not alter base line or standing PRA, aldosterone or proximal tubular function. Lithium does, however, induce an incomplete renal tubular acidosis.