This review investigates research evaluating the disinhibition hypothesis. This hypothesis postulates that in a sober state behavior is inhibited. When people are influenced by alcohol the inhibitions are supposed to be weakened and the motivating drives are postulated to become disinhibited and potent to influence behavior. This report reviews the effect of alcohol on nerve functions, on human sexuality, aggression, eating behavior, psychological conflicts, fluency in talk, social anxiety, violent crimes and the interaction of alcohol and social norms. It has been shown that individual subjective experiences sometimes indicate disinhibition (reduction of the forces holding back impulses) and objective behavior in some respects was different when the subject was intoxicated, but the mechanism that mediates behavior is not clear. It seems to be difficult to measure independently the forces restraining (inhibiting) the driving forces (uncontrolled impulses) at the same time as measuring these driving forces. The review concludes that there is no unambiguous support of the disinhibition hypothesis. An alternative hypothesis that seems to explain many behaviors in an inebriated individual is the 'time out' hypothesis which states that drunken behavior is influenced more by norms about what it should be than by the pharmacological effect of alcohol.