Results indicate that the non-homicide participants proved more problematic on many of the variables of analysis. They often began their delinquent careers earlier, had significantly greater numbers of total offenses, and had more violent offenses. The non-homicide participants often had less stable childhood histories, with more frequent placements out of the home and more frequent sibling delinquency. They also reported being more likely to use knives in crimes of violence than the homicide group. In measures of mental health and psychological functioning, the non-homicide participants reported greater problems with anger control, had less frequent positive early memories of their parents, and were more likely to endorse the belief that others were jealous of them, the dispositional factor that may bear some link with an antisocial orientation and violence link. The homicide participants were distinguishable on two key variables: they endorsed the greater availability of guns in homes and were more likely to report having taken guns from their homes in the past; and they reported a greater incidence of substance abuse at the time of their commitment offenses, suggesting that these factors operate as facilitating environmental influences and may be more significant than criminal histories and other personal history characteristics for understanding the dynamic context of the homicides. Data were collected in a secure detention program in Massachusetts where 33 incarcerated male juvenile offenders charged with or convicted of homicide were compared to 38 juvenile offenders charged with or adjudicated delinquent for violent but non-homicide offenses. Tables, references