Racial prejudice has been linked to influencing perceptions of criminal acts when the criminal is a member of a stigmatized group. The current study addresses reliance on stereotyping and emotional reactions to a crime when the participants are from an ethnically diverse samples (i.e., Latino and Asian Americans) and the crime is stereotypically congruent (i.e., African American assailant) or stereotypically incongruent (i.e., White assailant). Participants were videotaped reacting to a newspaper article about an assault and robbery crime in which the race of the suspect and victim were manipulated. Independent raters coded the reactions in terms of participant's emotional responses and the types of explanations given for the crime. Attributional styles for explaining criminal behavior are compared by suspect race (Black or White), victim ethnicity (Asian or Latino), and participant ethnicity (Asian or Latino). The fundamental assumptions underlining the "intergroup bias" and the "ultimate attribution error" were tested. Overall, the results did not fully support the current theoretical framework in that participants did not show a pronounced preference for in-group victims, or explain the perpetrators behavior using mainly dispositional attributions. Although participants used more situational than dispositional attributions overall to explain the perpetrators behavior, some evidence of activation of stereotypes was found. Qualitative analysis of attributions for criminal behavior differed between the Black suspect and the White suspect conditions. For instance, participants were more likely to attribute the Black perpetrators behavior being "gang-related" or a "rite of passage" but in the White suspect condition a common explanation was "childhood neglect" or "abuse." There was very little overlap in specific explanations given between conditions, which demonstrates that differential framework is being activated by the race of the suspect and that stereotypes may be influencing explanations. Results merit replication but demonstrate the importance of theory testing in diverse samples. Important ethnic group differences as well as implications, limitations of the current design, such as low variability in several of the judge-rated dependent variables, and future directions are discussed.