The examination of 27 murder cases revealed that for this subgroup of violent white-collar criminals, the threat of having their fraud detected is a motive to kill the person who threatens to reveal their criminal activities. This subgroup of killers is referred to in this article as “red-collar” criminals, because they straddle both the white-collar crime classification and the violent-crime classification. Both the descriptive homicide data and the literature review that revealed this subclassification of killers support three overriding impressions: red-collar criminals have the state of mind to do whatever it takes, including murder, in order to prevent someone from exposing their crimes; the victim of a red-collar killing may not be a co-perpetrator in the fraud, but is in a position to pose a threat to the continuation of a lucrative fraud scheme; and red-collar criminals have a history of antisocial and psychopathic tendencies. The existence of this subclassification of homicides suggests that forensic accountants and fraud examiners should be members of homicide investigation teams, in order to uncover the motive for murder by red-collar criminals. This study examined murder cases in which fraud pre-dated the murder. Information from each case was broken down into identifiable variables, such as type of fraud, age and gender of the defendant, and the defendant’s relationship to the victim. The proposed new crime classification, fraud-detection homicide, is compared to the existing FBI homicide classification in order to illustrate why fraud-detection homicide should be considered a separate classification. 42 references