Alcohol exerts numerous pharmacological effects through its interaction with various neurotransmitters and neuromodulators. Among the latter, the endogenous opioids play a key role in the rewarding (addictive) properties of ethanol. Three types of opioid receptors (mu, delta and kappa) represent the respective targets of the major opioid peptides (beta-endorphin, enkephalins and dynorphins, respectively). The rewarding (reinforcing) properties of mu- and delta-receptor ligands are brought by activation of the mesolimbic dopamine system which ascends from the ventral tegmentum of the midbrain (VTA) to rostral structures; of these, the nucleus accumbens (NAC) is of particular importance in drug addiction. In contrast, dysphoria results from activation of kappa-receptors. The neurochemical manifestations of these opposing effects are, respectively, increases and decreases in dopamine release in the NAC. Several lines of evidence indicate that alcohol interferes with endogenous opioid mechanisms which are closely linked with dopamine transmission in the mesolimbic pathway. The view that condensation products of dopamine and alcohol-derived aldehyde (tetrahydroisoquinolines) play a role remains controversial. There is, however, much information on the direct (acute and chronic) effects of alcohol on the binding properties of opioid receptors, as well as modulation of opioid peptide synthesis and secretion (e.g. a suggested increase in beta-endorphin release). In view of the reinforcing properties of alcohol, it is relevant to consider behavioural studies involving alcohol self-administration in rodents and primates. Low doses of morphine have been found to increase, and higher doses of the opiate to decrease, alcohol consumption. Conversely, opioid antagonists such as naloxone and naltrexone (which bind to non-selectively opioid receptors) have been shown to decrease alcohol consumption under various experimental conditions. Similar results have been reported when selective mu- or delta-receptor antagonists are administered. Results obtained in genetic models of high preference for alcohol also support the view that alcohol intake depends on the activity of the endogenous opioid reward system and that alcohol consumption may serve to compensate for inherent deficits in this system. One hypothetical model proposes that reward results from activation of mu-opioid receptors in the VTA and/or delta-receptor in the NAC; both these nuclei are targets of endogenous beta-endorphin. It is suggested that alcohol interferes with this reward pathway either directly or indirectly. The available experimental data accord well with those obtained from clinical studies which opioid antagonists have been used to prevent relapse in alcoholics. Conceptual considerations concerning communalities between various forms of addictions are also discussed in this review.