Neurotensin microinjection into the medullary nucleus raphe magnus (RMg) has been shown to both inhibit and facilitate the spinal nociceptive tail-flick reflex in a dose-dependent manner. Our study was designed to determine a potential involvement of spinal cholecystokinin octapeptide (CCK) in mediating neurotensin hyperalgesia from the RMg. Microinjection of neurotensin (50 ng) into the RMg of awake rats produced a facilitation of the tail-flick reflex that was completely inhibited by intrathecal (i.t.) administration of the nonselective CCK receptor antagonist proglumide (100 ng). Conversely, injection of a greater dose of neurotensin (5 micrograms) into the RMg produced an inhibition of the tail-flick reflex that was enhanced by i.t. proglumide. Intrathecal administration of the selective CCKB receptor antagonist L-365260 dose-dependently inhibited neurotensin hyperalgesia from the RMg (ID50 = 0.42 ng) at doses approximately 1000-fold less than that observed with the selective CCKA receptor antagonist devazepide (ID50 = 646 ng). Injection of CCK alone i.t. produced a biphasic response on the tail-flick reflex as lesser doses (0.1-0.3 ng) inhibited the reflex although greater doses (30-100 ng) facilitated it. Similar to supraspinal neurotensin hyperalgesia, the hyperalgesia observed with i.t. CCK (30 ng) was inhibited by i.t. L-365260 (ID50 = 0.59 ng) at doses approximately 1000-fold less than that observed with i.t. devazepide (ID50 = 630 ng). These data indicate that spinal CCK can both inhibit and facilitate spinal nociceptive responses. The facilitation of nociception observed with spinal CCK appears to involve CCKB receptors, which is consistent with the data in our study suggesting that spinal CCKB receptors mediate neurotensin hyperalgesia from the RMg via descending neuronal projections.