The effects of Native American race, crime severity, and intoxication on 293 college students' judgments of guilt were investigated. Participants read vignettes of a robbery or a robbery resulting in murder committed by a male defendant with a Native American or a White European surname. The defendant was described as highly intoxicated in half of the conditions. A 2 × 2 × 2 ANOVA revealed a marginally significant 3-way interaction. In the robbery–murder condition, the intoxicated Native American defendant was judged as guiltier than the Native American defendant with no intoxication information. We labeled this a stereotypical drunken-Indian bias. The results suggested that judgments given by lower and higher prejudiced participants canceled each other out.