The potential effect of intracerebroventricular (icv) alpha N-acetyl human beta-endorphin-(1-31) on morphine dependence was examined in mice and rats. Animals were rendered tolerant-dependent by subcutaneous (sc) implantation of an oily suspension (10 ml/Kg mouse and 3 ml/Kg rat) containing 0.1 g/ml of morphine. After 72 h of chronic morphine, 1 mg/Kg sc naloxone precipitated in both species a withdrawal syndrome that was moderate in animals pretreated with the acetylated derivative of beta-endorphin. Doses of 28 fmols/rat or 80 fmols/mouse alpha N-acetyl human beta-endorphin-(1-31) reduced the number of animals presenting the jumping behaviour, as well as the number of jumps recorded. Moreover, less than half of the rats presented the other withdrawal signs evaluated: squeak on touch, diarrhoea, chattering, chewing, ptosis and body shakes. This activity could be observed when alpha N-acetyl human beta-endorphin was injected 1 h to 24 h before naloxone; longer intervals resulted in a significant loss of this activity. The alpha 2 agonist clonidine given icv at pmol-nmol doses decreased the incidence of morphine withdrawal syndrome. Combinations of these two substances generally did not produce any further enhancement of the effects of clonidine and alpha N-acetyl beta-endorphin when used alone. Icv injections of the antagonist of alpha 2-adrenoceptors yohimbine prevented both clonidine and alpha N-acetyl beta-endorphin-(1-31) from reducing the jumping behaviour displayed by morphine-abstinent mice. It is suggested that alpha N-acetyl beta-endorphin produces this alleviation of the morphine withdrawal syndrome by improving the efficiency of alpha 2-mediated agonist effects after acting on a neural substrate that is distinct from the mu opioid receptor binding site.