Alcohol consumption and addiction have been related to anxiety and the anxiolytic effect of ethanol. It has been shown in mice that losers with repeated experience of social defeats are more anxious than winners with repeated experience of victories. Mice with a different social status were tested for their oral ethanol consumption using a free two bottle choice paradigm and for their social approach behaviour after ethanol consumption using the partition test, in which anxiety is an important component. In addition, the sensitivity of the animals for the kappa-opioid receptor agonist U-50,488H (2.5 mg/kg, s.c.) was assessed using the partition test, in which this drug has been shown to induce anxiolytic-like effects. Further, the effect of daily treatment with U-50,488H for 8 days on ethanol consumption was tested in animals that had consumed ethanol and were subjected during these 8 days to a period of 5 days of interruption of ethanol supply and subsequently to a period of 3 days of renewed access to ethanol. Losers consumed more ethanol than winners. Consumption of ethanol was accompanied by a decrease of anxiety level, as evidenced by an increased approach behaviour in the partition test. U-50,488H stimulated ethanol consumption after a period of 5 days of interruption of ethanol supply and drug treatment in the losers, but not in the winners. U-50,488H increased approach behaviour in the losers not consuming ethanol and decreased this behaviour in the winners, especially in those that had consumed ethanol. It is postulated that U-50,488H acts as a partial agonist in this respect. The increased anxiety may be related to the enhanced ethanol consumption in the losers, which may be of relevance for the etiology of alcohol addiction.