It has been suggested that morphine has dual effects; emetic effects and anti-emetic effects. The chemoreceptor trigger zone, which is outside the BBB, mediates the emetic effect. In contrast, the vomiting center mediates the anti-emetic effect of opioids. Thus, naloxone methiodide, which does not cross the BBB, antagonizes emetic effects of opioids. We studied whether naloxone methiodide alters abnormal motility pattern induced by morphine in gastrointestinal (GI) tract. Strain gauge force transducers were sutured on the serosal surface of upper GI tract to record the circular muscle contractions in eight dogs. The ventricular access system was implanted to inject morphine intracerebroventricularly (icv). Effects of icv-injection of morphine (0.3-3.0 mug/kg, bolus) on GI motility were studied during intravenous infusion of naloxone hydrochloride or naloxone methiodide. Icv-injection of morphine (3.0 mug/kg) induced retching and vomiting in all dogs tested. Phasic contractions of the jejunum were observed after icv-injection of morphine. These contractions in the jejunum migrated orally to the antrum (retrograde peristaltic contractions; RPCs). Both naloxone hydrochloride and naloxone methiodide treatment virtually abolished the emetic effects of morphine. Naloxone hydrochloride completely abolished morphine-induced RPCs in all dogs, whereas naloxone methiodide converted morphine-induced RPCs to anterograde peristaltic contractions (APCs) in 6 of 8 dogs. Our current study suggests that central opioids may induce APCs and prevent emesis in conscious dogs. Naloxone methiodide may be useful to prevent the undesired side effects of morphine.