In several pain models, tricyclic antidepressants (TCAs) have been shown to reduce nociception. In the present study, we evaluated the antinociceptive effect of metapramine (META) in 4 nociception tests: (1) the hot plate test; (2) the phenylbenzoquinone-induced writhing; (3) the tail flick test; and (4) the test of electrical stimulation of the tail. We further analysed, using META and clomipramine (CLOM), the eventual role of endogenous opioids in analgesia induced by TCAs. The analgesic effects of META and CLOM in the hot plate test and in the test of electrical stimulation of the tail were reversed by naloxone. On the other hand, we failed to demonstrate a potentiation of META- or CLOM-induced analgesia by acetorphan, an inhibitor of 'enkephalinase.' We also failed to show a potentiation of Met5-enkephalin intracerebroventricularly injected by the two TCAs. Moreover, the administration of the enzymatic inhibitor or of Met5-enkephalin led to a slight decrease of the analgesic effect of the TCAs. These results (1) indicate that in our 4 pain tests, META clearly reduces nociception and (2) provide evidence that the involvement of endogenous enkephalins in the analgesia induced by TCAs is improbable.