Studies of violence in stalking have treated interpersonal violence as a homogeneous phenomenon. This study was conducted to ascertain whether the associations of serious violence in stalking are the same as those of general violence in stalking. Of 85 stalkers referred to a forensic service, those who had committed acts of serious violence (homicide and serious assaults) were compared with those who had not on preselected clinical, demographic, and criminological variables. Associations of serious violence were found to differ from those reported for general violence. In particular, serious violence was significantly associated with an absence of criminal convictions and the presence of employment. There was no association with substance abuse, previous convictions for violence, or personality disorder. Different degrees of violence have different associations. This has implications for the development of violence prediction instruments and for violence prevention in stalking.