Neurotensin induced significant antinociceptive activity as measured in a variety of nociceptive tests 10 and 30 min following intracerebroventricular (i.c.v.) injection in mice. The lowest effective peptide doses were 25 ng in the writhing test, 25-50 ng in the tail-flick test, 50-100 ng in the hot-plate test and 2000 ng in the tail electrical stimulation test. The neurotensin related hexapeptide neuromedin N also displayed antinociceptive properties but only in the writhing and tail-flick tests. Furthermore, as compared to neurotensin, the neuromedin effects required higher doses. ED(50)'s for neurotensin and neuromedin in the writhing test were 70 ng and 1070 ng, respectively. Separate or combined injections of the endopeptidase 24.11 (enkephalinase) inhibitor thiorphan (l0?g) and the aminopeptidase inhibitor bestatin (50?g) did not affect tail-flick latencies. In contrast, i.c.v. injection of thiorphan together with an ineffective dose of neurotensin (25 ng) resulted in a significant antinociceptive effect. Bestatin did not modify tail-flick latencies in neurotensin-treated mice whether in the absence or presence of thiorphan. On the contrary, each of these peptidase inhibitors promoted antinociceptive effects of subthreshold doses of neuromedin (l?g) in the tail-flick test. Maximal antinociception was obtained by combining both inhibitors, thus conferring antinociceptive effects to neuromedin doses that were as low as 10 ng. Naloxone (0.5-2 mg/kg, s.c.) did not significantly reduced the antinociceptive effects of combinations of neurotensin and thiorphan and of neuromedin, thiorphan and bestatin. The data show that both neurotensin and neuromedin elicit analgesia in mice through an opiate independent mechanism. Furthermore, like enkephalin, neuromedin is readily degraded by brain endopeptidase 24.11 and bestatin sensitive aminopeptidase(s), whereas the resistance of neurotensin to aminopeptidase attack confers to this peptide a broader spectrum and longer duration of action than its congener neuromedin.