1. Intravenous administration of 1.0 mg/kg of morphine produces inhibition of the nociceptive tail-flick (TF) reflex, hypotension, and bradycardia in the pentobarbital-anesthetized rat. The present experiments examined peripheral, spinal, and supraspinal relays for inhibition of the TF reflex and cardiovascular responses produced by morphine (1.0 mg/kg iv) in the pentobarbital-anesthetized rat using 1) bilateral cervical vagotomy, 2) spinal cold block or mechanical lesions of the dorsolateral funiculi (DLFs), or 3) nonselective local anesthesia or soma-selective lesions of specific CNS regions. Intravenous morphine-induced inhibition of responses of unidentified, ascending, and spinothalamic tract (STT) lumbosacral spinal dorsal horn neurons to noxious heating of the hindpaw were also examined in intact and bilateral cervical vagotomized rats. 2. Bilateral cervical vagotomy significantly attenuated inhibition of the TF reflex and bradycardia produced by intravenous administration of morphine. Bilateral cervical vagotomy changed the normal depressor response produced by morphine into a sustained pressor response. Inhibition of the TF reflex in intact rats was not due to changes in tail temperature. 3. Spinal cold block significantly attenuated inhibition of the TF reflex, the depressor response, and the bradycardia produced by intravenous administration of morphine. However, bilateral mechanical transections of the DLFs failed to significantly affect either inhibition of the TF reflex or cardiovascular responses produced by this dose of intravenous morphine. 4. Microinjection of either lidocaine or ibotenic acid into the nuclei tracti solitarii (NTS), rostromedial medulla (RMM), or ventrolateral pontine tegmentum (VLPT) attenuated morphine-induced inhibition of the TF reflex. Similar microinjections into either the periaqueductal gray (PAG) or the dorsolateral pons (DLP) failed to affect morphine-induced inhibition of the TF reflex. 5. Microinjection of either lidocaine or ibotenic acid into the NTS, RMM, VLPT, DLP, or rostral ventrolateral medulla (RVLM) attenuated the depressor response produced by morphine, although baseline arterial blood pressure (ABP) was affected by ibotenic acid microinjections in the DLP. In all these cases, the microinjections failed to reveal a sustained pressor response as was observed with bilateral cervical vagotomy. Similar microinjections into the PAG failed to affect the depressor response produced by morphine. 6. The lidocaine and ibotenic acid microinjection treatments also showed that the bradycardic response produced by morphine depends on the integrity of the NTS, RMM, RVLM, and possibly the DLP, but not the PAG or VLPT. 7. Intravenous administration of morphine resulted in inhibition of responses of STT, ascending, or unidentified lumbosacral spinal dorsal horn neurons to noxious heating of the hind paw within 0.16 min of morphine administration. The inhibition produced by morphine was attenuated in rats receiving bilateral cervical vagotomy before morphine administration. Cumulative sum analyses indicated that inhibition of spinal nociceptive transmission started immediately after drug administration. All neurons studied had receptive fields on the glabrous skin of the plantar surface of the ipsilateral hind paw and responded to mechanical stimuli of both low and high intensity as well as noxious thermal stimulation (class 2 neurons). 8. These results demonstrate that inhibition of the TF reflex and spinal nociceptive transmission produced by intravenous administration of 1.0 mg/kg of morphine initially depends on the integrity of vagal afferents and several CNS relays. The time course and extent of vagal mediation of inhibition of spinal nociceptive transmission produced by intravenous morphine is virtually identical to that of the TF reflex. The depressor and bradycardic responses produced by morphine also depend on several CNS relays.