The literature presents too many false positives and false negatives to provide any support for predictors of violent juvenile behavior. There is no evidence of a separate category of 'violent juvenile offenders' either for research or for programs of rehabilitation or punishment. There are juveniles who occasionally commit violent acts, but they cannot be reliably identified. Violence by juveniles is apparently random and situational. Research on the causes and prediction of juvenile deviant behavior should continue but not under the assumption that there is a significant number of juveniles who consistently and predictably behave violently. The literature on violent juvenile behavior lacks discussion of adult roles in influencing violent juvenile behavior. Some of the literature suggests that a family's 'criminality' influences juvenile delinquency among the children and that older youth or adults force juveniles into gangs that model violent behavior, but there is little discussion in the literature of gang membership or adult leadership in gangs as variables useful for predicting juvenile violence. 14 notes.