The literature on the interaction of ethanol and stress is reviewed. Stress has amethystic properties in both experimental animals and human subjects when analgesia or various behavioral parameters were evaluated. In addition studies have shown that ethanol counteracts some effects of stress in non-alcoholic subjects. Improvement in performance and in the effective state of humans and reversal of some behavioral and pathological concomitants of stress in experimental animals have been reported. Although there is indication that ethanol improves the affective state in humans, reduction of anxiety has not been a universal finding. Recent studies have pointed out a number of variables (the drinking environment, cognitive set, mood and personality of subjects, prior experience with ethanol, sex, dose and type of beverage) which can significantly alter the effects of alcohol in human subjects. These may account for the variability in results in the literature. Although some studies have shown that alcohol ingestion increases under stressful conditions, others have failed to do so in both experimental animals and in humans. In alcoholics ethanol ingestion, in general, does not appear to relieve anxiety. In fact anxiety usually increases with time during a drinking binge. Therefore more research needs to be done to assess the validity of the anxiety-reducing theory for alcohol abuse. Possible mechanisms for the interaction of stress and alcohol are discussed.