The actions and toxicity of sodium nitroprusside (Na2Fe(CN)5 NO2H20) have been investigated periodically during the past 42 years without, however, definite establishment of the mechanisms on which the actions depend, or indications of its possible usefulness. The original claims of L. Herrmann, of his pupil Davidsohn, and of Arntz and Cromme, that the actions of nitroprusside depended upon liberated hydrocyanic acid suggested the possible use of the compound as a respiratory stimulant. The actions of the compound, therefore, have been re-investigated along different lines with the following results. The marked symptoms claimed by previous investigators, namely sustained, intense respiratory stimulation, followed sometimes by tremors and convulsions in all species, and accompanied by nausea and emesis in mammals and birds, have been confirmed. The odor of cyanogen post mortem, following large fatal doses only, has been confirmed, but chemical tests for cyanogen in the blood and tissues have proved inconclusive. Other tests were applied because small and non-fatal doses (2 mgm. per kilo or less hypodermically and about 0.1 to 0.2 mgm. per kilo intravenously), still gave profound effects without demonstrable cyanogen. These effects in narcotized and operated animals consisted of marked respiratory stimulation and a marked and sustained fall of blood pressure accompanied by increases in volume of peripheral organs and diminution in cardiac volume and contractions with increase or without change in rate. The circulatory actions were practically identical with those of sodium nitrite, but opposite to those of cyanide, in the same animals. A nitrite-type of action was further indicated from the facts that the vasomotor center still responded to reflex and asphyxial stimulations, the smooth muscle and sympathetic endings of the blood vessels were responsive in the usual way to epinephrine and the smooth muscle of excised organs was typically depressed. These actions might be expected from the nitroso (NO) group in the nitroprusside, and the respiratory stimulation would result from the depressor action, rather than from the alleged liberation of cyanogen.