It has been suggested that the endogenous opioid peptides, methionine and leucine enkephalin, participate only in naloxone-facilitated antinociceptive responses. To reassess this proposal, analgesic effects resulting from complete inhibition of enkephalin metabolism by intracerebroventricular (i.c.v.) administration of the mixed inhibitor RB 38A (R,S)HONHCOCH2CH(CH2 phi)CONHCH(CH2 phi)COOH) were compared to the effects of morphine (i.c.v.) in various assays commonly used to select analgesics: mouse hot plate-test, tail flick test with mice and rats, electrical stimulation of the tail (TES), paw pressure test with rats, and phenylbenzoquinone-induced writhing test with mice. The ED50s of morphine vs. ED50s of RB 38A in the writhing, hot plate (jumping) and tail flick tests with mice were 0.24 nmol vs. 38 nmol, 1 nmol vs. 36 nmol and 3.2 nmol vs. 285 nmol, respectively. RB 38A (ED30 153 nmol) was only 15 times less active in the tail flick test with rats than morphine and only halve as active in the paw pressure test. Noxious TES in rat was very sensitive to the inhibitory action of endogenous opioids protected by RB 38A, particularly the post-vocalization response which was also shown to be alleviated by antidepressants. All the analgesic effects observed were reversed by naloxone. This first direct evidence of analgesia resulting from peptidase inhibition, in the tail flick test with mice and rats, hot plate (paw lick) and TES shows that the pain suppressive effects of endogenous opioid peptides are not restricted to naloxone-facilitated noxious stimuli but occur more generally, in all morphine-sensitive tests. The differential effects of RB 38A in the various assays is likely to be related to the amount of enkephalins released and to the efficiency of peptidase inactivation in particular brain regions implicated in the control of a given nociceptive input. This mechanism could account for the reduction in side-effects compared to those of morphine following chronic administration of RB 38A.