Current thinking emphasizes that protracted small afferent input can evoke mechanisms that mediate a significant potentiation of spinal nociceptive processing and that this facilitory component has a unique pharmacology. To investigate the behavioral parallels of this spinal facilitation, we evaluated the effects of pre- and post-treatment of intrathecal morphine (mu agonist) and MK801 (N-methyl-D-aspartate [NMDA] antagonist) on the formalin test. Intraplantar formalin resulted in a biphasic appearance of flinching behavior (phase 1 = 0-5 min; phase 2 = 10-60 min). Morphine and MK801 were administered intrathecally 15 min before formalin injection in the pretreatment study and 9 min after formalin injection in the posttreatment study. Pretreatment with intrathecal morphine produced comparable dose-dependent suppressions of the phase 1 and phase 2 behaviors (ED50 = 0.5 micrograms [95% CI = 0.3-0.9] and 0.3 micrograms [95% CI = 0.1-0.7], respectively). Posttreatment with morphine also resulted in comparable suppression of the phase 2 response (ED50 = 0.2 micrograms [95% CI = 0.1-0.3]). At the highest dose of intrathecal morphine (10 micrograms), an almost complete suppression of formalin-evoked behavior was observed. Pretreatment with MK801 inhibited the second-phase response more strongly than the first-phase response (ED50 = 1.6 micrograms [95% CI = 0.5-5.7] vs. 0.1 microgram [95% CI = 0.3 - 0.4], respectively). In contrast, posttreatment with the highest dose of MK801 had no effect on the phase 2 response.(ABSTRACT TRUNCATED AT 250 WORDS)