A new study in Annals of Epidemiology categorizes homicides by victim count, revealing distinct characteristics that can help inform prevention strategies

September 2, 2026 – An updated analysis of US homicide data has found that intimate partner violence (IPV) often precipitates mass homicide and is also frequently associated with multi- and single-victim deaths. The study categorizes homicides into single-victim, multi-victim (2–3 victims), and mass (4 or more victims, excluding the perpetrator) fatalities, thereby revealing distinct demographics and circumstances. The findings of the study in Annals of Epidemiology, published by Elsevier, can help inform the development of tailored strategies to prevent homicide.

In the United States, homicides continue to be a public health problem. Overall, US homicide rates have fluctuated in recent years, but little research has examined patterns and variation across homicide types.

This study analyzed data from 115,873 homicide deaths in the CDC’s National Violent Death Reporting System (NVDRS) from 2018 to 2023 to examine patterns across homicide types. The vast majority (90.5%; 104,858) were single-victim cases, however, almost one in twelve victims died in multi-victim homicides, claiming 10,123 lives across 4,789 events (8.7%). This is a higher percentage of multi-victim homicide deaths than found in the prior study using NVDRS data from 2003 to 2017. The higher proportion of multi-victim homicide deaths may reflect expanded geographic coverage of NVDRS states over time as more states joined the system, or potential changes in homicide patterns at the national level, as evidenced by increases in national homicide rates during the corresponding time period.

Mass homicides involving four or more casualties accounted for less than 1.0% of the total, resulting in 892 deaths across 178 incidents, with individual cases spanning from 4 to 51 fatalities. Mass and multi-victim homicides frequently occurred in homes, while single homicides most often occurred in public locations like streets or vehicles.

Demographic patterns, however, were consistent with previous findings. Mass homicides more often involved non-Hispanic (NH) White victims and children, whereas single- and multi-victim homicides more often involved NH Black victims and acquaintances or friends as suspects.

Key findings

• Sex: Both victims and suspects were primarily male across all categories.
• Age: Almost half of victims in multi-victim and single-victim homicides were aged 18–34 years. In mass homicides, 29.8% of victims fell in this age group.
• Victim demographics: The highest proportion of mass homicide victims were NH White (39.7%); the highest proportion of multi- and single- homicide victims were NH Black (multi, 47.8%; single, 54.8%).
• Suspect demographics: The highest proportion of mass homicide suspects were NH White (32.2%) and NH Black (30.1%). Among multi-victim and single-victim homicides, the highest proportion of suspects were NH Black (37.6% and 36.8%, respectively).
• Firearms: Firearms were the primary method of injury used across all homicide types (mass, 75.8%; multi, 86.3%; single, 76.4%).
• Precipitating circumstances: IPV was a common precipitating circumstance in nearly one-third of mass homicides (31.0%) and was frequently associated with multi- (16.9%) and single- (15.0%) victim homicides as well. Family relationship problems and a recent or impending crisis within the past two weeks were common circumstances associated with mass and multi-victim homicides. For multi-victim and single-victim homicides, arguments and criminal activity were common precipitators.

The findings underscore that the differences among these homicide types are key to developing tailored prevention strategies.

Patrick S. Sullivan, DVM, PhD, Emory University, Editor-in-Chief of Annals of Epidemiology, adds, "This analysis is a reminder of what public data systems make possible. By linking death certificates, medical examiner findings, and law enforcement reports, the National Violent Death Reporting System lets us see that homicide is not one problem but several — each with its own affected populations and precipitating circumstances. The findings of this study offer important guideposts for the policies and interventions most likely to reduce these tragic and preventable outcomes."