1. The mechanism of catecholamine secretion after coronary occlusion was investigated in open-chest, anaesthetized dogs; the blood-bathed organ technique was used for continuous measurement of the changes in concentration of circulating catecholamines.
2. In dogs with increased output of adrenaline during the first hour after acute coronary occlusion, topical application of lignocaine to the infarcted area of the heart, spinal block at C1, bilateral section of thoracic splanchnic nerves, or ganglionic blockade virtually abolished adrenaline secretion.
3. Adrenaline secretion was abolished by bilateral vagotomy in 50% of the dogs, decreased in 28% and unchanged in the remaining 22%. Bretylium or guanethidine had no effect on adrenaline secretion in the early stages of myocardial infarction.
4. Seventeen of nineteen reserpinized dogs failed to secrete catecholamines during the first hour of coronary occlusion, even though the adrenal medulla responded to bradykinin, acetylcholine or nicotine. Noradrenaline infused either into the unoccluded carotid artery or intravenously restored adrenal medullary secretion.
5. When a sustained noradrenaline release occurred after coronary ligation, it was abolished by topical application of lignocaine to the ischaemic area of the heart or by ganglionic blockade. Neither bilateral vagotomy nor section of both thoracic splanchnic nerves suppressed noradrenaline liberation.
6. It is concluded that the adrenal medullary secretion of adrenaline which occurs in the early stages of myocardial infarction in the dog is induced reflexly from stimulation of cardiac receptors at the site and the boundary of the infarct. The reflex involves vagal as well as extra-vagal pathways and supraspinal structures. Enhanced liberation of noradrenaline in the early stage of infarction may reflect release from postganglionic sympathetic nerve endings in the heart.