Daily urinary excretion of catechol amines in normal rats and in rats from which the adrenal medullae had been removed has been determined by a photofluorimetric method. In both groups, reserpine (2 mg/kg, intraperitoneally) produces: (1) A decrease in the urinary excretion of noradrenaline which persists for more than 3 weeks; this action is not influenced by monoamine oxidase inhibitors and mecamylamine. (2) An increase, within 20 to 68 hr, in the urinary excretion of adrenaline, even though the urine of rats without adrenal medullae does not usually contain adrenaline. These effects are prevented by monoamine oxidase inhibitors and, in the normal animals, are reduced by mecamylamine. In both groups, dexamphetamine (6 mg/kg, intraperitoneally) produces an increase in the excretion of adrenaline and noradrenaline, the adrenaline appearing in the urine of the rats without adrenal medullae within 20 to 44 hr. Mecamylamine prevents the effect of dexamphetamine on the excretion of noradrenaline. Dexamphetamine, administered within a week of reserpine treatment, produces its usual effects on the urinary excretion of catechol amines in normal rats, but has no effect in rats without adrenal medullae. The results are discussed with regard to both the mechanism by which reserpine and dexamphetamine influence the peripheral stores of adrenaline and noradrenaline, and the significance of the adrenal and extra-adrenal chromaffin system.