We have tested the widely accepted hypothesis that digitalis, in nontoxic concentration, can act directly on cells of the sinoatrial pacemaker to cause slowing. The effect of autonomic nerves on sinoatrial cells was eliminated either by chronic cardiac denervation or by use of autonomic blocking agents such as atropine, 0.5 mg/kg, and propranolol, 2 mg/kg, or MJ-1999, 3 mg/kg. When the effects of autonomic nerves were eliminated, acetylstrophanthidin, 0.120 mg/kg, failed to cause slowing of sinus rate in the anesthetized cat and the anesthetized dog. Similarly, ouabain, 5 X 10~"M, and acetylstrophanthidin, 1.5 /ig/ml, after prior administration of autonomic blocking agents failed to cause slowing of the isolated rabbit right atrium perfused with Tyrode's solution. These results indicate that the sinus slowing caused by cardiac glycosides may be mediated entirely by alteration of the neural control of the heart and that the drug, in nontoxic concentration, does not have a direct negative chronotropic action on the mammalian sinoatrial node. has been presented by Mendez et al. (5) and by Nadeau and James (6). Prior to their studies (5, 6), any sinus bradycardia induced by cardiac glycosides and not blocked by vagotomy or by administration of atropine was attributed to a direct effect of the glycoside on the sinoatrial node (7,8). Therefore, at least three actions of cardiac glycosides which might cause sinus bradycardia have been described: (1) a vagal effect, (2) an effect which modifies the normal sympathetic influence on heart rate, hereafter referred to as "the sympathetic modifying effect," and (3) a direct negative chronotropic effect. The vagal and sympathetic modifying effects of cardiac glycosides alter sinus rate by modifying reflex mechanisms and are properly classified as indirect chronotropic effects. If,