In 2 laboratory studies, the authors tested the hypothesis that intoxicated risk taking results from alcohol's effects on negative outcome expectancies. Young adults (N = 107) consumed alcohol or no alcohol and made ratings of the likelihood that negative and positive consequences would result from a variety of risky activities. Consistent with study hypotheses, participants rated negative consequences as less likely when they were intoxicated than when sober. Results were replicated in a second study (N = 88), which further showed that alcohol, rather than expectancy set, contributed to these reduced perceptions of risk. Findings provide the first experimental evidence that alcohol intoxication may contribute to risk-taking behavior be altering expectations about negative consequences.