Three studies are reported. In study 1, vanadium concentration was estimated by neutron activation analysis in hair, whole blood, serum and urine from 13 patients suffering from depressive psychosis and then when on recovery- Vanadium concentration of hair, whole blood and serum decreased significantly with recovery, but there was no significant change in 24-h urinary excretion or in renal clearance of vanadium. In study 2, vanadium concentration was estimated by neutron activation analysis in serum and urine of 31 patients with depressive psychosis and of 27 normal controls. Mean renal clearance of vanadium was significantly lower and mean serum vanadium concentration significantly higher in depressed patients than in controls. Mean 24-h excretion of vanadium did not differ between the two groups. Vanadium excretion did not correlate with urine volume, with serum concentration or with age. In study 3, erythrocyte Na-K ATPase activity and serum vanadium concentrations were estimated in 58 patients. There was a strong negative correlation between the two, supporting the suggestion that changes in tissue vanadium concentration may explain the changes in sodium transport which occur in depressive psychosis.