This study investigates the effects of alcohol, expectancy, and alcohol-related beliefs on self-reported anxiety and self-disclosure behavior in a social interaction situation. Seventy-two female social drinkers were assigned to eight conditions in a 2 × 2 × 2 factorial balanced-placebo design, controlling for drink content, expectancy, and beliefs. Results show that alcohol expectancy resulted in an anxiety reduction in subjects who believed that alcohol has a positive influence on social behavior, whereas expectancy had no effect in subjects with negative alcohol beliefs. Furthermore, subjects who expected alcohol were less anxious when they had positive alcohol beliefs than when they had negative beliefs, the effect of alcohol expectancy on social anxiety in women in moderated by differences in the effect of alcohol expectancy on social anxiety in women is moderated by differences in the content of their alcohol-related beliefs. We conclude that the inconsistency in previous results about the effect of alcohol expectancy on social behavior might be explained by differences in subjects' alcohol beliefs. Finally, as neither of the three factors had any effect on self-disclosure behavior, we suggest that this behavioral measure is unrelated to self-reported anxiety.