1. Supraspinal substrates mediating vagal afferent stimulation (VAS)-induced inhibition of the nociceptive tail-flick reflex were examined by the use of the soma-selective neurotoxin ibotenic acid and the nonselective local anesthetic lidocaine. Fifty rats were studied in the lightly anesthetized state maintained with pentobarbital sodium. 2. The threshold intensity of VAS required to inhibit the tail-flick reflex to a cut-off latency of 10 s was established in all rats. Ibotenic acid (5 or 10 micrograms, 0.5 microliter) or lidocaine (4%, 0.5 microliter) was then microinjected into various regions of the brain stem followed by reestablishment of the intensity of VAS required to produce inhibition of the tail-flick reflex. 3. Microinjections of ibotenic acid into the ipsilateral nucleus tractus solitarius (NTS), medial rostroventral medulla (principally the nucleus raphe magnus; NRM), or bilaterally into the dorsolateral pons (principally the locus coeruleus/subcoeruleus; LC/SC), significantly increased the threshold intensity of VAS required to inhibit the tail-flick reflex. Microinjections of ibotenic acid into either the rostral or caudal ventrolateral medulla (RVLM or CVLM, respectively) ipsilateral to the vagus nerve stimulated or ipsilateral LC/SC did not significantly affect the inhibition produced by VAS. Arterial blood pressure decreases produced by VAS were significantly attenuated or eliminated after microinjections of ibotenic acid into the NTS, RVLM, CVLM, or NRM. Lidocaine microinjected into the ipsilateral CVLM also significantly increased the intensity of VAS required to inhibit the tail-flick reflex. 4. These outcomes obtained with behavioral measures are consistent with the outcomes of the preceding study using electrophysiological measures in establishing that cells in the NTS, LC/SC, and NRM regions and fibers of passage in the CVLM are important in mediating the inhibitory effects of VAS. The present studies confirm previous reports of the importance of the RVLM and CVLM in VAS-produced depressor responses but also demonstrate that the NRM is critical for this cardiovascular response.