Intracerebroventricularly injected morphine is 50-fold more potent in arresting intestinal peristalsis in rats, mice or guinea pigs than morphine administered systemically. Using quaternary naloxone as narcotic antagonist, it has been demonstrated that the peripheral pathway of the centrally mediated constipatory effect of morphine does not involve opioid peptidergic mechanisms. Further, this effect is not due to the release of opioid peptides from the pituitary, since hypophysectomy fails to affect the antipropulsive activity of morphine. On the other hand, the intestinal motility can be affected directly by activation of opiate receptors located in the gut. This was best demonstrated with loperamide, which exhibits predominantly a peripheral site of action. Thus, two mechanisms of the action of morphine on gastrointestinal propulsive activity have been demonstrated. One arises in the central nervous system (CNS) and is mediated peripherally not by opioid peptidergic pathways, whereas the other is due to a direct action of morphine on the gut.